DETAILED ACTION
This final Office action is responsive to Applicant’s amendment filed August 20, 2026. Claims 11-12 and 15 are canceled. Claims 1 and 13-14 have been amended. Claims 1-10, 13-14, and 16-22 are presented for examination.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant's arguments filed August 20, 2026 have been fully considered but they are not persuasive.
Regarding the rejection under 35 U.S.C. § 101, Applicant’s arguments are deemed to be persuasive. The Examiner agrees that the currently claimed details of training the models present more than generally applying machine learning to otherwise abstract ideas. Instead, the claims present machine learning (including neural network) details that go beyond simply being applied to the abstract ideas and these details present more than a generic, conventional use of technology. The rejection is withdrawn.
Regarding the art rejections, Applicant submits that Cohen does not address the claims, as amended, especially since Cohen performs post hoc analysis. The Angelov reference has been introduced into the rejection in order to help address the claim amendments.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-10, 13-14, and 16-22 are rejected under 35 U.S.C. 103 as being unpatentable over Cohen et al. (US 2024/0249299) in view of Yuan et al. (CN-112926615-A, referencing the English Translation) in view of Angelov et al. (Angelov, P. et al. "Towards Explainable Deep Neural Networks (xDNN)." Neural Networks 130 (2020) 185-194. Available online 11 July 2020.).
[Claim 1] Cohen discloses a method of generating a prediction information (abstract – “…applying an explainability model to the decision model to generate one or more predictors or drivers of the output of the decision model, wherein the one or more predictors or drivers (1) are features of the channel affinity sub-model and/or the content affinity sub-model and (2) provide an explanation of an effect of the action on the target variable.”), comprising:
acquiring feature data corresponding to a target object for a plurality of object demand influence features (¶ 21 – “Another aspect provides a system for enhancing explainability of one or more models that are useable to increase sales of one or more products. The system may comprise: one or more processors; and a memory storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: generating one or more predictive models based at least in part on (i) a set of target variables, (ii) a set of features, and (iii) a set of decision variables, wherein the features are predictive of and have an influence on the target variables…”; ¶¶ 44. 48. 61. 66, 88, 90, 93, 93, 113 – Learning and training are performed on various sets of data, including specific features (¶¶ 48, 61, 66).);
determining an object category corresponding to the target object according to the feature data (¶¶ 21, 60, 63, 88 – Any defining aspect of the analyzed product/object, such as a time period, a facility, constraints, conditions, a model used, etc. can be interpreted as an example of an object category.);
determining at least one information to be predicted for the target object according to the object category (abstract – “An example method may comprise (a) using a decision model to predict an action that a sales representative should take to maximize a target variable…”; ¶ 59 – “X may be features that are predictive or believed to be predictive of the target variable Y. With continued reference to a predictive model for a sales organization, X may include demographic information of about a customer (e.g., age, gender, educational background, and the like). The demographic profile of a customer may, for example, be predictive of the type of communication that the customer prefers to receive (e.g., a phone call rather than an email). X may also include data about the customer's business. For example, if the sales organization is a pharmaceutical sales organization and the customer is a health care provider (“HCP”), X may include data about the HCP's patient population. X may also include a history of previous contacts with the customer, including the substance, dates and times, and outcomes of in-person visits to the customer, emails sent to the customer, documents provided to the customer, webinars and conferences attended by the customer, and the like. X may be configured in multiple ways, depending on whether the prediction model is time-dependent or not.”; ¶ 87 – “The decision model is based on a predictive model f(x,d) that maps features, including facility visits, to sales. D*(x) may represent the constrained decision model.”; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales.”; ¶ 112 – “The recommendations may be actions that are predicted to maximize a target variable (e.g., likelihood of making a sale, or sales amount).”);
generating at least one first feature demand prediction information for a target time according to at least one first feature demand prediction model and the feature data, wherein the at least one first feature demand prediction model is in one-to-one correspondence with the at least one information to be predicted, and the first feature demand prediction model is an interpretable model (¶ 47 – “Explainability models may be models that are inherently interpretable or models that explain other uninterpretable models. Explainability models may include deep explanation models, interpretable models, and models of models (“model induction”). Deep explanation models are neural networks in which nodes are identified as features so that the weights of the various layers illuminate the drivers of the neural network. Interpretable models are models that are inherently interpretable, including linear models, parametric models, tree models, Bayesian models, and the like. And model induction is a technique whereby a more interpretable model is built on top of an underlying model. Examples of models that may be used in model induction are local interpretable model-agnostic explanations (LIME), Shapley additive explanations (SHAP), counterfactual local explanations via regression (CLEAR), Anchors, and leave one covariate out (LOCO).”; ¶ 87 – “A pharmaceutical company wants to determine the quantity of quarterly visits to each facility that the company serves (e.g., doctor's offices, clinics, and hospitals) that maximizes the sale of each of two therapeutic products. The company is motivated to reduce costly individual visits, potentially replacing them with group conferences or emails and freeing up resources so that more facilities can be served with the same resource overhead. However, in-person visits may result in more sales. The company builds a decision model that determines the number of visits to each facility that maximizes the sale of the two therapeutics, considering historical data. The decision model is based on a predictive model f(x,d) that maps features, including facility visits, to sales. D*(x) may represent the constrained decision model.”; ¶ 59 – “X may be features that are predictive or believed to be predictive of the target variable Y. With continued reference to a predictive model for a sales organization, X may include demographic information of about a customer (e.g., age, gender, educational background, and the like). The demographic profile of a customer may, for example, be predictive of the type of communication that the customer prefers to receive (e.g., a phone call rather than an email). X may also include data about the customer's business. For example, if the sales organization is a pharmaceutical sales organization and the customer is a health care provider (“HCP”), X may include data about the HCP's patient population. X may also include a history of previous contacts with the customer, including the substance, dates and times, and outcomes of in-person visits to the customer, emails sent to the customer, documents provided to the customer, webinars and conferences attended by the customer, and the like. X may be configured in multiple ways, depending on whether the prediction model is time-dependent or not.”; ¶ 63 – “The system 200 can also include a decision model generator 210. The decision model generator 210 can generate a decision model from the predictive model. The decision model can predict the values of decision variables D that maximize the target variable Y, where decision variables D are a subset of features X. Decision variables may be variables over which a person or entity has some control. For example, a sales representative can control the content and timing of emails, topics of discussion on a phone call, and the like. The predictive problem may therefore be recharacterized as f(X,D)=Y. The goal of finding f( ) may be to use the information contained therein to make decisions about what values of D maximize Y. This may be expressed as the unconstrained decision model:”; ¶ 75 – “The set of features may include features that are or are believed to be predictive of the target variable. The set of features may include decision variables. Decision variables may be actions that are under the control of and executed by the person or entity that implements or uses the predictive model (e.g., a sales representative). In other words, decision variables may be variables that can be deliberately controlled. The set of features may also include variables that cannot be controlled directly that are also predictive of the target variable. For example, a company's pre-existing market share, which the company may not be able to control directly, may be predictive of sales.”; ¶ 76 – “In the case of a pharmaceutical company, the set of features may include demographic data associated with an HCP. The demographic data may be predictive, for example, of whether the HCP will respond to a particular mode of contact but not another (e.g., a phone call, but not an email). The demographic data may include age, gender, education background, and the segment membership of the HCP. Additionally or alternatively, the set of features may include data that is indicative of the HCP's patient population (e.g., the percentage of the HCP's patient population that has a particular disease). Additionally or alternatively, the set of features may include a contact history associated with the HCP and sales representatives of the pharmaceutical company. The contact history may include one or more of the following: (1) a number of visits by the one or more sales representatives to the HCP, (2) topics of conversations during the visits, (3) a number of email correspondences sent by the one or more sales representatives to the HCP, (4) topics of the email correspondences sent, (5) documents relating to the pharmaceutical product provided by the one or more sales representatives to the HCP, (6) webinars attended by the one or more sales representatives and the HCP, and (7) conferences attended by the one or more sales representatives and the HCP. Such contact history and corresponding sales data may indicate which types of contact are most valuable to the pharmaceutical company.”).
Cohen discloses inputting the at least one first feature demand prediction information and the feature data into a pre-trained second feature demand prediction model, so as to generate at least one second feature demand prediction information and a total demand prediction information for the target time, wherein the second feature demand prediction model is an uninterpretable model (Cohen: fig. 3 –
PNG
media_image1.png
598
294
media_image1.png
Greyscale
; ¶¶ 61-63 – “[0061] The predictive model generator 205 can train the predictive model using a supervised, semi-supervised, or unsupervised learning process, for example. A supervised predictive model can be trained using labeled training inputs, i.e., features X and corresponding target variables Y. Features X can be provided to an untrained or partially trained version of the predictive model to generate a predicted output. The predicted output can be compared to the known target variable Y for that set of features X, and if there is a difference, the parameters of the predictive model can be updated. A semi-supervised predictive model can be trained using a large number of unlabeled features X and a small number of labeled features X. An unsupervised predictive model, e.g., a clustering or dimensionality reduction model, can find previously unknown patterns in features X. [0062] The predictive model generated by the predictive model generator 205 may be a neural network (e.g., a feedforward neural network, a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory network (LSTM), etc.), an autoencoder, a regression model, a decision tree, a random forest model, a support vector machines, a Bayesian network, a clustering model, a reinforcement learning algorithm, or the like. [0063] The system 200 can also include a decision model generator 210. The decision model generator 210 can generate a decision model from the predictive model. The decision model can predict the values of decision variables D that maximize the target variable Y, where decision variables D are a subset of features X. Decision variables may be variables over which a person or entity has some control. For example, a sales representative can control the content and timing of emails, topics of discussion on a phone call, and the like. The predictive problem may therefore be recharacterized as f(X,D)=Y. The goal of finding f( ) may be to use the information contained therein to make decisions about what values of D maximize Y. This may be expressed as the unconstrained decision model:”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 84 – “In a retail example, the set of features may include demographic and purchasing history associated with a particular customer. The features may be predictive as to whether the customer at a particular store, when the customer may make a purchase, what types of items the customer may purchase, or other target variables. The decision variables in such a scenario may be features under which retail companies or individual retail employees have some control, such as distribution of coupons and employee interactions with the customer. Thus, the decision model may determine the relative importances of the decision variables to a target outcome, while an explanation model may provide insight into how the decision variable features interact with one another.”; ¶ 93 – “FIG. 5 is a scatter plot of the predictive model's predicted values against the actual target values for each of the therapeutics. The plots show a strong diagonal pattern which confirms that the model fit is good. As described above, the approaches to building an explanation model evaluate the predictive model either on a sample of the data set used to train the model or on a set of counterfactuals. In this case, counterfactuals were used to generate observations that cover the complete space of the predictors. The system may use these data to build the decision model.”; ¶ 110 – “The system may build a linear model with the weighted hypercube values as predictors. To determine contributions of predictor variables (predictor), the system may test different model targets that are particular percentage deviations from an optimum value. The plot of FIG. 10C shows values of coefficients for two estimated LIME explanation models: for a constrained model and for an unconstrained model. The predictors are the horizontal axis and the values of their coefficients are on the vertical access. The table gives the exact values of the coefficients. The value of r2 for the unconstrained model is 0.97 and for the constrained model is 0.98, meaning that the model is an effective predictive tool. The plot shows that, for each model, the variables “appointment”, “conference”, and “visit” were highly determinative of the prediction. Although these results match those for the recursive model, the LIME model may not determine the multivariate impact of the predictors as explainers in the decision model.”; fig. 5, ¶ 94 – “The surface defined by the predictions of f( ) is an 8-dimensional surface. Since the observations that comprise the surface are from the predictions of a random forest model and not a parametric model, there are discontinuities in the surface, as the plots of FIG. 6 show. FIG. 6 shows the surface across 4-dimensions for two quarters. The surface varies across the quarters, across the facilities, and across the products. The first row in each plot shows data for product 1 and the second is for product 2; as the plots move from left to right the facility decile increases. Some of the variance and fluctuations in the plots is caused by the discontinuities of the random forest model and some is caused by the hidden variables that are not shown in the plots. FIG. 6 illustrates more detail on this prediction surface and provides insight into the decision model. The plot on the right has blue and red lines—these are the maximum and the 95% quantile for the prediction in each of the identified dimensions, respectively. The value of visit where the maximum intersects is the value for d*(x) for that set of predictors. Since there is variance associated with the predictive model method, the average number of visits where the prediction for those values is above the 95% quantile within that bin of predictors is used as the value for d*(x).”; ¶ 88 – “The company trained the predictive model on historical sales data of the two products to different medical facilities. The historical sales data included quarterly sales data for each facility for each of the two products. A particular data record contained an indication of the product (product), quarter (qtr), and facility of the data record; a code indicating the decile of the sales of the facility (facility); the number of scheduled visits sales representatives made to an HCP in the facility (appointment); the number of conferences that HCPs within the facility attended (conference): the number of group meetings that HCPs within the facility attended (group); the number of emails sent to HCPs within the facility (email); and the number of unscheduled visits to HCPs within the facility (visit).”; ¶ 47 – “Explainability models may be models that are inherently interpretable or models that explain other uninterpretable models. Explainability models may include deep explanation models, interpretable models, and models of models (“model induction”). Deep explanation models are neural networks in which nodes are identified as features so that the weights of the various layers illuminate the drivers of the neural network. Interpretable models are models that are inherently interpretable, including linear models, parametric models, tree models, Bayesian models, and the like. And model induction is a technique whereby a more interpretable model is built on top of an underlying model. Examples of models that may be used in model induction are local interpretable model-agnostic explanations (LIME), Shapley additive explanations (SHAP), counterfactual local explanations via regression (CLEAR), Anchors, and leave one covariate out (LOCO).”; ¶ 54 – “One framework for SHAP is additive feature attribution methods, which provides a representation of relative feature importances within a prediction model. Additive feature attribution may estimate an underlying prediction model as a sum of transformed, weighted feature terms. The method may determine the weights by minimizing a loss function. Features which are more heavily weighted may be thus inferred to be more important to the prediction.”).
Cohen does not explicitly disclose that the second feature demand prediction model is a deep learning neural network model, and a prediction process of the second feature demand prediction model depends on interpretable intermediate variables provided by the at least one first feature demand prediction information.
Angelov uses prototypes-based classifiers as part of a reasoning process as an alternative to post hoc analysis (Angelov: p. 186: col. 2 – “Prototypes-based classifiers are a reasoning process that do not consider post hoc analysis (Biehl et al., 2016).”). The prototypes-based classifier is used with transfer learning and an xDNN approach, which imparts some understandability to deep learning models, such as deep neural networks (Angelov: p. 186: cols. 1-2). Angelov further states, “The baseline approach (Zhao et al., 2020) is a Deep Neural Network approach which is ‘black box’ (offers no interpretability).” (Angelov: p. 192: col. 2)
While Angelov does not focus on features and models being related to demand predictions, Angelov does analogously evaluate relationships among features and provide explainability for the types of networks generally used by Cohen, such as a convolutional neural network (CNN) (Cohen: ¶ 62).
Angelov discloses the following steps:
generating at least one first feature information for a target time according to at least one first feature model and the feature data, wherein the at least one first feature model is in one-to-one correspondence with the at least one information to be predicted, and the first feature model is an interpretable model (Angelov: p. 187: cols. 1-2 – “1. Features layer: (Defines the data space) The Feature Layer is the first phase of the proposed xDNN method. This layer is in charge of extracting global features vector from the images. This first layer can be formed by more traditional ‘handcrafted’ methods such as GIST (Solmaz et al., 2013) or HoG (Mizuno et al., 2012). Alternatively, it can be formed by the fully connected layer (FCL) of the pre-trained convolutional neural network approaches such as AlexNet (Krizhevsky et al., 2012), VGG–VD–16 (Simonyan & Zisserman, 2014), and Inception (Szegedy et al., 2015), residual neural networks such as Resnet (He et al., 2016) or Inception-Resnet (Szegedy et al., 2017), etc. Using pre-trained deep neural network approach allows automatic extraction of more abstract and discriminative high level features. In this paper, pre-trained VGG–VD–16 DCNN is employed for feature extraction. According to Ren et al. (2016), VGG–VD–16 has a simple structure and it can achieve a better performance in comparison with other pre-trained deep neural networks. Initialization: Meta-parameters for the xDNN are initialized with the first observed data sample (image). The proposed algorithm works per class; therefore, all the calculations are done for each class separately. P ←1; µ←xi; layer from VGG–VD–16 provides a 1 × 4096 dimensional vector. (3) where µ denotes the global mean of data samples of the given class. P is the total number of the identified prototypes from the observed data samples (images). “; p. 190: col. 1 – Images may be grouped (and totaled) by implied time periods (such as day or night) and/or weather conditions, as seen in Section 4.1.);
inputting the at least one first feature information and the feature data into a pre-trained second feature model, so as to generate at least one second feature information and a total for the target time, wherein the second feature model is an uninterpretable model, the second feature model is a deep learning neural network model, and a process of the second feature model depends on interpretable intermediate variables provided by the at least one first feature demand prediction information (Angelov: p. 186: col. 2 – “Prototypes-based classifiers are a reasoning process that do not consider post hoc analysis (Biehl et al., 2016).”). The prototypes-based classifier is used as a part of transfer learning and an xDNN approach, which imparts some understandability to deep learning models, such as deep neural networks (Angelov: p. 186: cols. 1-2). Angelov further states, “The baseline approach (Zhao et al., 2020) is a Deep Neural Network approach which is ‘black box’ (offers no interpretability).” (Angelov: p. 192: col. 2); Figs. 2, 3 --
PNG
media_image2.png
546
806
media_image2.png
Greyscale
; p. 188: cols. 1-2 – “4. Prototypes layer: The prototypes identification layer is the core of the proposed xDNN classifier. This layer is responsible to provide the clearly explainable model. The xDNN classifier is free from prior assumptions about the data distribution type, as well as the random or deterministic nature of the data. In contrast, it empirically extracts the distribution from the data samples (images) bottom up (Angelov & Gu, 2019). The prototypes are independent from each other. Therefore, one can change the structure by adding a new prototype without influencing the other already existing prototypes. In other words, the proposed xDNN is highly parallelizable and suitable for evolving form of application where new prototypes may be added (if the data pattern requires this). The proposed xDNN method is trained per class forming a set of prototypes per class. Therefore, all the calculations are done for each class separately. Prototypes are the local peaks of the data density (and typicality) identified in the previous layers/stages of the algorithm from the images of the corresponding class based on their feature vectors.”; p. 190: col. 1 – Images may be grouped (and totaled) by implied time periods (such as day or night) and/or weather conditions, as seen in Section 4.1.).
Cohen specifically uses neural networks to predict sales (i.e., demand), as seen in ¶¶ 44, 46, 60, 62-63 of Cohen. Angelov presents an approach to providing some explainability for a deep learning neural network.
The Examiner submits that it would have been obvious to one of ordinary skill in the art before the effective filing date of Applicant’s invention to modify Cohen to fully perform the steps of:
generating at least one first feature demand prediction information for a target time according to at least one first feature demand prediction model and the feature data, wherein the at least one first feature demand prediction model is in one-to-one correspondence with the at least one information to be predicted, and the first feature demand prediction model is an interpretable model;
inputting the at least one first feature demand prediction information and the feature data into a pre-trained second feature demand prediction model, so as to generate at least one second feature demand prediction information and a total demand prediction information for the target time, wherein the second feature demand prediction model is an uninterpretable model, the second feature demand prediction model is a deep learning neural network model, and a prediction process of the second feature demand prediction model depends on interpretable intermediate variables provided by the at least one first feature demand prediction information
because “the proposed approach [proposed in Angelov] can obtain higher performance using less training data due to its prototype-based nature. Experiments have demonstrated that the proposed xDNN approach is able to produce highly accurate results surpassing state-of-the-art methods for different challenging datasets. Moreover, xDNN presents highly interpretable results that can be presented in the form of IF ...THEN logical rules, Voronoi tessellations, and/or typicality (empirically derived form of pdf) in a closed analytical form allowing further analysis. Because of its recursive, non-iterative and non-parametric form it allows computationally very efficient implementations to be realized.” (Angelov: p. 193: col. 1) Additionally, Angelov explains, “The proposed xDNN approach requires very little computational resources (no need for GPUs) and short training times (in the order of seconds). The proposed approach, xDNN is prototype-based. Prototypes are actual training data samples (images), which have local peaks of the empirical data distribution called typicality as well as of the data density. This generative model is identified in a closed form and equates to the pdf but is derived automatically and entirely from the training data with no user- or problem-specific thresholds, parameters or intervention. The proposed xDNN offers a new deep learning architecture that combines reasoning and learning in a synergy. It is non-iterative and non-parametric, which explains its efficiency in terms of time and computational resources. From the user perspective, the proposed approach is clearly understandable to human users. Results for some well-known benchmark data sets such as iRoads, Caltech-256, Caltech-101, and COVID CT-scan show that xDNN outperforms the other methods including state-of-the-art deep learning approaches in terms of accuracy, time to train and offers an explainable classifier.” (Angelov: p. 193: cols. 1-2)
Cohen does not explicitly disclose performing replenishment processing on the target object according to the at least one second feature demand prediction information and the total demand prediction information.
Yuan uses long-tail data for an article of interest to predict future sales of the article (Yuan: pp. 9-10 – “Specifically, after determining the sale of the article, can be based on the sale of each article, responding to the service request of the main server, so that the main server receives the response of each slave server, integrating, obtaining the sale of all product terminal request, feeding back to the terminal; so as to make the related personnel to prepare goods according to the sale of the product.
wherein, if the original service data comprises the business data of the product with high sales volume, the main server can predict the future sales according to the existing prediction mode, also can be distributed to each slave server for prediction, specifically can be set according to the actual requirement, the invention is not limited.
The processing method of the service provided by this embodiment, after receiving the service request, according to the service request comprises the service data to determine the target long-tail data, and according to the target long-tail data and pre-obtained diagonal matrix to determine the target long-tail data corresponding to the correlation matrix; so as to decompose the correlation matrix based on the characteristic decomposition, obtaining the normalized characteristic vector corresponding to the correlation matrix, and determining the correlation random number corresponding to the target long-tail data based on the pre-obtained diagonal matrix and the normalized feature vector, based on the correlation random data; determining the sale of each article included in the target long-tail data; because of non-zero filling the zero data in the long-tail data, for predicting the future sale of the article, increasing the information amount contained in the long-tail data, improving the accuracy of the product future sales prediction, so as to reduce the storage cost.”).
Additionally, Yuan explains that the prediction of future sales volume for a product is for product restocking (Yuan: p. 6 – “The processing method of service provided by the embodiment of the invention is suitable for but not limited to the application scene for determining future sales volume for product restocking.”).
The Examiner submits that it would have been obvious to one of ordinary skill in the art before the effective filing date of Applicant’s invention to modify Cohen to incorporate the step of performing replenishment processing on the target object according to the at least one second feature demand prediction information and the total demand prediction information in order to “increas[e] the information amount contained in the long-tail data, improving the accuracy of the product future sales prediction, so as to reduce the storage cost” (as suggested on p. 10 of Yuan), which would specifically be useful in more efficient and accurate replenishment planning for products that often take longer to sell relative to other products.
[Claim 2] Cohen does not explicitly disclose wherein determining at least one information to be predicted for the target object according to the object category comprises:
determining, in response to determining that the object category is a long-tail object category, a demand trend feature information as an information to be predicted.
Yuan uses long-tail data for an article of interest to predict future sales of the article (Yuan: pp. 9-10 – “Specifically, after determining the sale of the article, can be based on the sale of each article, responding to the service request of the main server, so that the main server receives the response of each slave server, integrating, obtaining the sale of all product terminal request, feeding back to the terminal; so as to make the related personnel to prepare goods according to the sale of the product.
wherein, if the original service data comprises the business data of the product with high sales volume, the main server can predict the future sales according to the existing prediction mode, also can be distributed to each slave server for prediction, specifically can be set according to the actual requirement, the invention is not limited.
The processing method of the service provided by this embodiment, after receiving the service request, according to the service request comprises the service data to determine the target long-tail data, and according to the target long-tail data and pre-obtained diagonal matrix to determine the target long-tail data corresponding to the correlation matrix; so as to decompose the correlation matrix based on the characteristic decomposition, obtaining the normalized characteristic vector corresponding to the correlation matrix, and determining the correlation random number corresponding to the target long-tail data based on the pre-obtained diagonal matrix and the normalized feature vector, based on the correlation random data; determining the sale of each article included in the target long-tail data; because of non-zero filling the zero data in the long-tail data, for predicting the future sale of the article, increasing the information amount contained in the long-tail data, improving the accuracy of the product future sales prediction, so as to reduce the storage cost.”).
Additionally, Yuan explains that the prediction of future sales volume for a product is for product restocking (Yuan: p. 6 – “The processing method of service provided by the embodiment of the invention is suitable for but not limited to the application scene for determining future sales volume for product restocking.”).
The Examiner submits that it would have been obvious to one of ordinary skill in the art before the effective filing date of Applicant’s invention to modify Cohen wherein determining at least one information to be predicted for the target object according to the object category comprises:
determining, in response to determining that the object category is a long-tail object category, a demand trend feature information as an information to be predicted
in order to “increas[e] the information amount contained in the long-tail data, improving the accuracy of the product future sales prediction, so as to reduce the storage cost” (as suggested on p. 10 of Yuan), which would specifically be useful in more efficient and accurate replenishment planning for products that often take longer to sell relative to other products.
[Claim 3] Cohen discloses wherein determining at least one information to be predicted for the target object according to the object category comprises:
determining, in response to determining that the object category is a first object category, a similar object demand prediction information as an information to be predicted, wherein an object corresponding to the first object category has no value transfer data (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.” Having “no value transfer data” may mean that there are no constraints, for example. Additionally, having “no value transfer data” could mean that an object is new (like a new product), which Cohen addresses in ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…”).
[Claim 4] Cohen discloses wherein determining at least one information to be predicted for the target object according to the object category comprises:
determining, in response to determining that the object category is a second object category, a sensitive information corresponding to the target object, wherein value transfer data of an object corresponding to the second object category meets a preset transfer condition (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.”; ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…” Having a “present transformation condition” may refer to conditions and constraints that must be met and/or are at least evaluated.); and
determining, in response to determining that the sensitive information indicates that the target object is an object of which a value attribute transformation meets a preset transformation condition, a first value related feature influence information and a demand trend feature information as an information to be predicted respectively (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.”; ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…” Having a “present transformation condition” may refer to conditions and constraints that must be met and/or are at least evaluated.; ¶ 75 – “The set of features may include features that are or are believed to be predictive of the target variable. The set of features may include decision variables. Decision variables may be actions that are under the control of and executed by the person or entity that implements or uses the predictive model (e.g., a sales representative). In other words, decision variables may be variables that can be deliberately controlled. The set of features may also include variables that cannot be controlled directly that are also predictive of the target variable. For example, a company's pre-existing market share, which the company may not be able to control directly, may be predictive of sales.”; ¶ 76 – “In the case of a pharmaceutical company, the set of features may include demographic data associated with an HCP. The demographic data may be predictive, for example, of whether the HCP will respond to a particular mode of contact but not another (e.g., a phone call, but not an email). The demographic data may include age, gender, education background, and the segment membership of the HCP. Additionally or alternatively, the set of features may include data that is indicative of the HCP's patient population (e.g., the percentage of the HCP's patient population that has a particular disease). Additionally or alternatively, the set of features may include a contact history associated with the HCP and sales representatives of the pharmaceutical company. The contact history may include one or more of the following: (1) a number of visits by the one or more sales representatives to the HCP, (2) topics of conversations during the visits, (3) a number of email correspondences sent by the one or more sales representatives to the HCP, (4) topics of the email correspondences sent, (5) documents relating to the pharmaceutical product provided by the one or more sales representatives to the HCP, (6) webinars attended by the one or more sales representatives and the HCP, and (7) conferences attended by the one or more sales representatives and the HCP. Such contact history and corresponding sales data may indicate which types of contact are most valuable to the pharmaceutical company.”).
[Claim 5] Cohen discloses wherein after determining, in response to determining that the sensitive information indicates that the target object is an object of which a value attribute transformation meets a preset transformation condition, a first value related feature influence information and a demand trend feature information as an information to be predicted respectively, the method further comprises:
determining, in response to determining that the sensitive information indicates that an association degree between value transfer data corresponding to the target object and an object flow information meets a preset association condition, a second value related feature influence information and the demand trend feature information as an information to be predicted respectively (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.”; ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…” Having a “present transformation condition” may refer to conditions and constraints that must be met and/or are at least evaluated.; ¶ 75 – “The set of features may include features that are or are believed to be predictive of the target variable. The set of features may include decision variables. Decision variables may be actions that are under the control of and executed by the person or entity that implements or uses the predictive model (e.g., a sales representative). In other words, decision variables may be variables that can be deliberately controlled. The set of features may also include variables that cannot be controlled directly that are also predictive of the target variable. For example, a company's pre-existing market share, which the company may not be able to control directly, may be predictive of sales.”; ¶ 76 – “In the case of a pharmaceutical company, the set of features may include demographic data associated with an HCP. The demographic data may be predictive, for example, of whether the HCP will respond to a particular mode of contact but not another (e.g., a phone call, but not an email). The demographic data may include age, gender, education background, and the segment membership of the HCP. Additionally or alternatively, the set of features may include data that is indicative of the HCP's patient population (e.g., the percentage of the HCP's patient population that has a particular disease). Additionally or alternatively, the set of features may include a contact history associated with the HCP and sales representatives of the pharmaceutical company. The contact history may include one or more of the following: (1) a number of visits by the one or more sales representatives to the HCP, (2) topics of conversations during the visits, (3) a number of email correspondences sent by the one or more sales representatives to the HCP, (4) topics of the email correspondences sent, (5) documents relating to the pharmaceutical product provided by the one or more sales representatives to the HCP, (6) webinars attended by the one or more sales representatives and the HCP, and (7) conferences attended by the one or more sales representatives and the HCP. Such contact history and corresponding sales data may indicate which types of contact are most valuable to the pharmaceutical company.”).
[Claim 6] Cohen discloses wherein determining at least one information to be predicted for the target object according to the object category comprises:
determining, in response to determining that the object category is a seasonal object category, a sensitive information corresponding to the target object (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.”; ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…” Having a “present transformation condition” may refer to conditions and constraints that must be met and/or are at least evaluated.; ¶ 75 – “The set of features may include features that are or are believed to be predictive of the target variable. The set of features may include decision variables. Decision variables may be actions that are under the control of and executed by the person or entity that implements or uses the predictive model (e.g., a sales representative). In other words, decision variables may be variables that can be deliberately controlled. The set of features may also include variables that cannot be controlled directly that are also predictive of the target variable. For example, a company's pre-existing market share, which the company may not be able to control directly, may be predictive of sales.”; ¶ 76 – “In the case of a pharmaceutical company, the set of features may include demographic data associated with an HCP. The demographic data may be predictive, for example, of whether the HCP will respond to a particular mode of contact but not another (e.g., a phone call, but not an email). The demographic data may include age, gender, education background, and the segment membership of the HCP. Additionally or alternatively, the set of features may include data that is indicative of the HCP's patient population (e.g., the percentage of the HCP's patient population that has a particular disease). Additionally or alternatively, the set of features may include a contact history associated with the HCP and sales representatives of the pharmaceutical company. The contact history may include one or more of the following: (1) a number of visits by the one or more sales representatives to the HCP, (2) topics of conversations during the visits, (3) a number of email correspondences sent by the one or more sales representatives to the HCP, (4) topics of the email correspondences sent, (5) documents relating to the pharmaceutical product provided by the one or more sales representatives to the HCP, (6) webinars attended by the one or more sales representatives and the HCP, and (7) conferences attended by the one or more sales representatives and the HCP. Such contact history and corresponding sales data may indicate which types of contact are most valuable to the pharmaceutical company.”; ¶¶ 65, 78 – Seasonality is taken into account.); and
determining, in response to determining that the sensitive information indicates that the target object is an object of which a value attribute transformation meets a preset transformation condition, a first value related feature influence information, a demand trend feature information and a seasonal feature influence information as an information to be predicted respectively (¶ 17 – “In some embodiments, the explanation model may comprise a global explanation model. Alternatively, the global explanation model may comprise an unconstrained global decision tree. In some embodiments, the global explanation model may comprise a constrained global decision tree.”; ¶ 82 – “In some cases, the system may apply the explainability modeling (e.g., recursive partitioning) over the entire set of features used to train the decision model, resulting in a global explanation model (e.g., global decision tree). The global explanation model may be a constrained global explanation model in that it considers constraints applied to the decision model, or it may be an unconstrained global explanation model. In other cases, however, the system may apply explainability modeling over only a subset of the features used to train the decision model, e.g., a margin of the space instead of the entire space, resulting in a local explanation model. In the case of recursive partitioning, for example, this may result in a local decision tree.”; ¶ 44 – “The present disclosure provides methods for explaining models that drive decision-making processes for businesses. Such models may be referred to as “decision models” in this disclosure. A decision model may include a predictive model, e.g., a machine learning (ML) model, that is trained on historical data and limited by one or more constraints and identifies decisions that optimize some business financial objective. The constraints may be operational constraints imposed on the business that limit the range of practical outputs that the predictive model can generate. Additionally or alternatively, the constraints may be rules set by the business that align with the goals of the business that likewise limit the range of decision outputs that the predictive model can generate and which optimizes the business objective. The trained decision model can determine one or more optimal actions for maximizing one or more target variables. The target variables may be business metrics, e.g., sales metrics. The methods described herein can comprise generating an explanation model from the decision model. The explanation model may be useable to gain insight into the structure and function of the model.”; ¶ 60 – “These decision variables may be variables on which humans may have control, and thus may allow humans to calibrate or optimize their actions (e.g., contacts from pharma reps to HCPs) to achieve desired results (e.g., increased sales or prescriptions filled). The values of decision variables that achieve desired results may not be feasible in the real world. Further, entities (e.g., businesses or regulators) may bar persons from taking actions represented by decision variables. In these situations, the system may add constraints to the decision model to better simulate real-world conditions or reflect real-world needs.”; ¶ 65 – “In practice, brand management and sales operations teams may also specify certain rules. Such rules may result from various plans and goals that may not be captured in the relationship between (X,D) and Y. For example, a brand team may want to prioritize the sale of a new product on the marketplace. Additionally or alternatively, the brand team may specify rules for interacting with uncontrolled publications, rules that require visits when commercial metrics change in statistically relevant ways, rules for timing interactions with seasonal commercial drivers, rules for coordinating messaging across products brands, and the like. Let R denote the set of rules and D denote the union of constraints and rules, namely D=C U R…” Having a “present transformation condition” may refer to conditions and constraints that must be met and/or are at least evaluated.; ¶ 75 – “The set of features may include features that are or are believed to be predictive of the target variable. The set of features may include decision variables. Decision variables may be actions that are under the control of and executed by the person or entity that implements or uses the predictive model (e.g., a sales representative). In other words, decision variables may be variables that can be deliberately controlled. The set of features may also include variables that cannot be controlled directly that are also predictive of the target variable. For example, a company's pre-existing market share, which the company may not be able to control directly, may be predictive of sales.”; ¶ 76 – “In the case of a pharmaceutical company, the set of features may include demographic data associated with an HCP. The demographic data may be predictive, for example, of whether the HCP will respond to a particular mode of contact but not another (e.g., a phone call, but not an email). The demographic data may include age, gender, education background, and the segment membership of the HCP. Additionally or alternatively, the set of features may include data that is indicative of the HCP's patient population (e.g., the percentage of the HCP's patient population that has a particular disease). Additionally or alternatively, the set of features may include a contact history associated with the HCP and sales representatives of the pharmaceutical company. The contact history may include one or more of the following: (1) a number of visits by the one or more sales representatives to the HCP, (2) topics of conversations during the visits, (3) a number of email correspondences sent by the one or more sales representatives to the HCP, (4) topics of the email correspondences sent, (5) documents relating to the pharmaceutical product provided by the one or more sales representatives to the HCP, (6) webinars attended by the one or more sales representatives and the HCP, and (7) conferences attended by the one or more sales representatives and the HCP. Such contact history and corresponding sales data may indicate which types of contact are most valuable to the pharmaceutical company.”; ¶¶ 65, 78 – Seasonality is taken into account.).
[Claim 7] Cohen discloses wherein generating at least one first feature demand prediction information for a target time according to at least one first feature demand prediction model and the feature data comprises:
for each information to be predicted among the at least one information to be predicted, performing a first input step (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110), comprising:
determining, in response to determining that the information to be predicted is a demand trend feature information, demand trend feature data corresponding to the demand trend feature information among the feature data (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”);
determining a first feature demand prediction model corresponding to the demand trend feature information as a demand trend information prediction model (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”); and
inputting the demand trend feature data into the demand trend information prediction model pre-trained, so as to output a demand trend prediction information as a first feature demand prediction information for the target time (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”).
[Claim 8] Cohen discloses for each information to be predicted among the at least one information to be predicted, performing a second input step, comprising:
determining, in response to determining that the information to be predicted is a first value related feature influence information, first value related feature data corresponding to a first value related feature among the feature data (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”);
determining a first feature demand prediction model corresponding to the first value related feature influence information as a first demand information prediction model (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”); and
inputting the demand trend prediction information and the first value related feature data into the first demand information prediction model pre-trained, so as to output a first demand prediction information under influence of the first value related feature as a first feature demand prediction information for the target time (Figs. 3, 7-10C; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”).
[Claim 9] Cohen discloses for each information to be predicted among the at least one information to be predicted, performing a second input step, comprising:
determining, in response to determining that the information to be predicted is a second value related feature influence information, second value related feature data corresponding to a second value related feature among the feature data (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”);
determining a first feature demand prediction model corresponding to the second value related feature influence information as a second demand information prediction model (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”); and
inputting the demand trend prediction information and the second value related feature data into the second demand information prediction model pre-trained, so as to output a second demand prediction information under influence of the second value related feature as a first feature demand prediction information for the target time (Figs. 3, 7-10C; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”).
[Claim 10] Cohen discloses for each information to be predicted among the at least one information to be predicted, performing a third input step, comprising:
determining, in response to determining that the information to be predicted is a seasonal feature influence information, seasonal feature data corresponding to a seasonal feature among the feature data (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”);
determining a first feature demand prediction model corresponding to the seasonal feature influence information as a third demand information prediction model (Fig. 3; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”); and
inputting the seasonal feature data and the demand trend prediction information into the third demand information prediction model pre-trained, so as to output a third demand prediction information under influence of the seasonal feature as a first feature demand prediction information for the target time (Figs. 3, 7-10C; ¶¶ 17, 44, 47, 54, 60-63, 65, 75-76, 78, 82, 84, 88, 93-94, 110; ¶ 96 – “In the plots on the left, the estimate lines show the number of visits that maximize sales as a function of facility sales size, number of emails sent, and number of appointments. Appointments increase for the plots further to the right and emails sent increase for the plots further toward the top. The plots show that the value of visits increases with facility size when there are fewer appointments (estimate lines with positive slopes on the left), but that trend inverts as appointments grow (kernel estimator lines with negative slopes on the right). One might expect that appointments are more important as facilities grow. The plots on the left also show that the impact of the number of emails sent is more subtle (only a small variation in the slope of the estimate lines in the same column).”; ¶¶ 65, 78 – Seasonality is taken into account.; ¶ 107 – “The bar charts in FIG. 10A show the coefficient values for the sampled instances. Positive values are interpreted as increases in the predictor driving increases in the number of visits that optimize sales. Notice that in the observations, increasing the quarter is associated with increasing visits in the sales-optimizing scenarios. This is consistent with the observations from the recursive partitioning explanation models as shown, for example, in FIGS. 8 and 9.”; ¶ 108 – “The plot shows the strong influence of the quarter on the optimal number of visits for maximizing sales. What cannot be seen in this plot is the detail that the recursive partitioning reveals in, for example, FIG. 9, where for smaller facilities it is favorable to have fewer appointments in the second half of the year in comparison to larger facilities, where it is favorable to have more appointments in the first half of the year.”).
[Claims 13, 16-22] Claims 13 and 16-22 recite limitations already addressed by the rejections of claims 1-8 above; therefore, the same rejections apply.
Furthermore, Cohen discloses an electronic device, comprising: one or more processors; and a storage device configured to store one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to at least perform the disclosed operations (Cohen: ¶¶ 21-24).
[Claim 14] Claim 14 recites limitations already addressed by the rejection of claim 1 above; therefore, the same rejection applies.
Furthermore, Cohen discloses a non-transitory computer readable medium storing a computer program, wherein the computer program, when executed by a processor, at least performs the disclosed operations (Cohen: ¶¶ 21-24).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Chakraborty, D. et al. "Interpretable vs. Noninterpretable Machine Learning Models for Data-Driven Hydro-Climatological Process Modeling." – Compares interpretable vs. noninterpretable machine learning models.
Yip et al. (US 2024/0379237) – Feeds data output from models which may be interpretable or non-interpretable into a model that may be interpretable or non-interpretable (Fig. 2A).
Bhatia et al. (US 2021/0365770) – Uses boosting to train models sequentially (¶ 52).
Dalli et al. (US 2021/0232915) – Maps the output of convoluted features back into the original input in a CNN-XNN prediction network (¶ 63).
Dalli et al. (US 2022/0138532) – Uses transfer learning to make a neural network interpretable (¶ 101).
Schwinn et al. (US 2025/0033213) – Determines an intermediate result that may be interpretable or uninterpretable (¶¶ 13, 17).
Cintas et al. (US 2023/0281492) – Discusses grey-box and white-box access to an AI model (¶ 59).
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SUSANNA M DIAZ whose telephone number is (571)272-6733. The examiner can normally be reached M-F, 8 am-4:30 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Brian Epstein can be reached at (571) 270-5389. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/SUSANNA M. DIAZ/
Primary Examiner
Art Unit 3625A
Prosecution Insights