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 . Claims 1-20 are presented in the case.
Specification
The title of the invention is not descriptive. A new title is required that is clearly indicative of the invention to which the claims are directed. The following title is suggested: “Stochastic Model-Based Abnormality Detection”
Information Disclosure Statement
The information disclosure statements submitted on 06/14/2023 and 01/26/2026 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Priority
Acknowledgment is made of applicant's claim for foreign priority based on application JP2022-096452 filed in Japan on 06/15/2022. Copies are available online.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”)
Claim 1, 19 and 20 have the following abstract idea analysis.
Step 1: The claim is directed to “a method, system and crm”. The claims are directed to the statutory categories accordingly.
Step 2A Prong 1: claims recite the abstract idea limitations of "calculate an estimate indicating a degree of being in the specific state based on a distribution of a plurality of the output values output from the data analysis model for the plurality of times". These limitations include mathematical concepts see MPEP § 2106.04(a)(2)) where it cites "performing a resampled statistical analysis to generate a resampled distribution". The specification also provides example calculation of estimate including for abnormalities and based on distribution See USPGPUB ¶34. Thus, these steps are an abstract idea in the “mathematical concept”. Other sections of the claims such as "input the input data to the data analysis model repeatedly a plurality of times", and "store a data analysis model trained in advance by using training data configured to output an output value indicating whether a target to be analyzed is in a specific state in response to the data analysis model receives input data on the target to be analyzed, the data analysis model including a parameter that includes a random variable;" are advanced processes, too generic or high level to be listed as a judicial exception given the available descriptions and MPEP comparisons.
Step 2A Prong 2: The judicial exceptions recited in these claims are not integrated into a practical application. Merely invoking "a data analysis model", "input data", "a processor", "storage" or "memory" does not yield eligibility. Claims are still in line with mental concepts such as claim 1, 19 and 20 are not specific to a practical application. The additional elements as such are processors and instructions which do not include specialized hardware. See MPEP § 2106.05(f).
Claim 1, 19 and 20 do not include a particular field but even doing so may not be sufficient to overcome the abstract idea rejection. Merely applying an model to a field or data without an advancement in the new field or new hardware is ineligible. MPEP § 2106.05(h).
Step 2B: The claims do not contain significantly more than their judicial exceptions. Processors, memory and other hardware are in their standard forms in the field. These additional elements are well-understood, routine, and conventional activity, see MPEP 2106.05(d)(II). Claims lacks any particular "how" or algorithm for a solution in a field in a novel way. Claims require more specificity on processes that would be incapable of simple mathematics, mental processes or use more substantial structure than conventional devices such as non-textbook implementations.
Regarding claims 2-18 merely narrow the previously recited abstract idea limitations with more abstract concepts and/or routine fundamental processes. For the reasons described above with respect to claim 1 this judicial exception is not meaningfully integrated into a practical application, or significantly more than the abstract idea. Abstract idea steps 1, 2A prong 1 and 2 remain the same as independent analysis above. See specification for more practical application concepts as none are seen in claims 2-18.
With respect to step 2B These claims disclose similar limitations described for the independent claims above and do not provide anything significantly more than mathematical or mental concepts. Claims 2-18 recite the additional elements of "acquire control data of a motor control apparatus for a motor, the motor being configured to drive a mechanism of an industrial machine, input, as the input data, the control data to the data analysis model repeatedly a plurality of times, and calculate, as the specific state, an estimate indicating a degree of occurrence of an abnormal phenomenon specific. the memory is configured to store a plurality of the data analysis models configured to output, respectively, a plurality of the output values indicating whether a plurality of the abnormal phenomena that are mutually different occur, and the circuitry is configured to calculate an estimate indicating a degree of occurrence of each of the plurality of the abnormal phenomena that are mutually different based on a plurality of the distributions of, respectively, a plurality of output values, the plurality of output values being obtained by the input unit inputting the control data to each of the plurality of the data analysis models repeatedly a plurality of times. identify a unit phenomenon that has occurred in the mechanism based on the control data, and input unit phenomenon data related to the unit phenomenon to the data analysis model repeatedly a plurality of times. identify a unit phenomenon that has occurred in the mechanism based on the control data, and input unit phenomenon data related to the unit phenomenon to the data analysis model repeatedly a plurality of times. wherein the circuitry is configured to calculate the estimate in such a manner that the degree of being in the specific state becomes lower as the distribution varies. wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. input the training data to the data analysis model that is trained repeatedly a plurality of times; and determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. input the training data to the data analysis model that is trained repeatedly a plurality of times; and determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. input the training data to the data analysis model that is trained repeatedly a plurality of times; and determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and determine a largest candidate out of the candidates as the threshold. input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and determine a largest candidate out of the candidates as the threshold. Input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and determine a largest candidate out of the candidates as the threshold. wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values within a predetermined range in the distribution. calculate an indicator related to a variation in the distribution and a mean of a plurality of the output values; and calculate the estimate based on the indicator and the mean. wherein the circuitry is configured to calculate, as the estimate, a value indicating a probability that a specific abnormal phenomenon has occurred in the target to be analyzed based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution." These elements are more abstract concepts, generic applications to a field of use or well-understood, routine, conventional activity (see MPEP § 2106.05(d) and can't be simply appended to qualify as significantly more or being a practical application. What type of application, or structure of components beyond generic machine learning is still unknown for these claims. Claims 3-6 Motor related data is used an a different type of data and not integrated as a practical application. Therefore claims 2-18 also recite abstract ideas that do not integrate into a practical application or amount to significantly more than the judicial exception, and are rejected under U.S.C. 101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 of this title, 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 and 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh et al. (US 20220366223 A1 hereinafter Fathallah-Shaykh) in view of Simonis et al. (US 20220101666 A1 hereinafter Simons)
As to independent claim 1, Fathallah-Shaykh teaches a data analysis system comprising: [system ¶5]
a memory [memory ¶5] configured to store a data analysis model trained in advance by using training data configured to output an output value indicating whether a target to be analyzed is in a specific state in response to the data analysis model receives input data on the target to be analyzed, the data analysis model including a parameter that includes a random variable; and [neural network model that classifies into states (disease or not) (training data and input ¶4), ¶11 "disease classification, object detection"], [random variables ¶12 "Bayesian framework, model parameters, i.e., the weights and biases, are defined as random variables"]
circuitry configured to: [circuit ¶32]
calculate an estimate indicating a degree of being in the specific state based on a distribution of a plurality of the output values output from the data analysis model for the plurality of times. [calculates confidence and matrices based on a distribution and outputs (degree of being in a state (the prediction)) ¶5 " approximate the mean and covariance of each respective tensor normal distribution passing through the non-linear activation function of each non-linear perceptron"…" compute a confidence in the prediction based at least in part on the mean matrix and the covariance matrix of the output vector."]
Fathallah-Shaykh does not specifically teach input the input data to the data analysis model repeatedly a plurality of times;
However, Simons teaches input the input data to the data analysis model repeatedly a plurality of times; and [continual monitoring and evaluating ¶29, ¶64, deploys models regularly ¶17 "model parameters of the fault classification model trained in this way can then be transmitted to the motor vehicle once or at regular intervals." ]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the classification modeling disclosed by Fathallah-Shaykh by incorporating the input the input data to the data analysis model repeatedly a plurality of times disclosed by Simons because both techniques address the same field of machine learning and by incorporating Simons into Fathallah-Shaykh enables more statistically significant predictions with models[Simons ¶31, ¶62]
As to dependent claim 17, the rejection of claim 1 is incorporated, Fathallah-Shaykh and Simons further teach wherein the circuitry is configured to: calculate an indicator related to a variation in the distribution and a mean of a plurality of the output values; and [Fathallah-Shaykh mean matrix and covariance ¶5]
calculate the estimate based on the indicator and the mean. [Fathallah-Shaykh mean matrix and covariance used in confidence calculation ¶5 " compute a confidence in the prediction based at least in part on the mean matrix and the covariance matrix of the output vector."]
As to dependent claim 18, the rejection of claim 1 is incorporated, Fathallah-Shaykh and Simons further teach wherein the circuitry is configured to calculate, as the estimate, a value indicating a probability that a specific abnormal phenomenon has occurred in the target to be analyzed based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. [Simonis probability of a fault ¶29 "determine a probability of the existence of an actual fault and from this to derive measures as appropriate and/or to determine a residual service life of a relevant component."], [Simonis anomaly degree threshold ¶64]
As to independent claim 19, Fathallah-Shaykh teaches a data analysis method comprising: [¶5]
data analysis model being trained in advance by using training data configured to output an output value indicating whether a target to be analyzed is in a specific state in response to the data analysis model receives the input data on the target to be analyzed, the data analysis model including a parameter that includes a random variable; and [neural network model that classifies into states (disease or not) (training data and input ¶4), ¶11 "disease classification, object detection"], [random variables ¶12 "Bayesian framework, model parameters, i.e., the weights and biases, are defined as random variables"]
calculate an estimate indicating a degree of being in the specific state based on a distribution of a plurality of the output values output from the data analysis model for the plurality of times. [calculates confidence and matrices based on a distribution and outputs (degree of being in a state (the prediction)) ¶5 " approximate the mean and covariance of each respective tensor normal distribution passing through the non-linear activation function of each non-linear perceptron"…" compute a confidence in the prediction based at least in part on the mean matrix and the covariance matrix of the output vector."]
Fathallah-Shaykh does not specifically teach input the input data to the data analysis model repeatedly a plurality of times;
However, Simons teaches input the input data to the data analysis model repeatedly a plurality of times; and [continual monitoring and evaluating ¶29, ¶64, deploys models regularly ¶17 "model parameters of the fault classification model trained in this way can then be transmitted to the motor vehicle once or at regular intervals." ]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the classification modeling disclosed by Fathallah-Shaykh by incorporating the input the input data to the data analysis model repeatedly a plurality of times disclosed by Simons because both techniques address the same field of machine learning and by incorporating Simons into Fathallah-Shaykh enables more statistically significant predictions with models[Simons ¶31, ¶62]
As to independent claim 20, Fathallah-Shaykh teaches non-transitory computer-readable recording medium containing a program for causing circuitry to implement processing, the processing comprising: [memory, processor and instructions ¶5]
data analysis model being trained in advance by using training data configured to output an output value indicating whether a target to be analyzed is in a specific state in response to the data analysis model receives the input data on the target to be analyzed, the data analysis model including a parameter that includes a random variable; and [neural network model that classifies into states (disease or not) (training data and input ¶4), ¶11 "disease classification, object detection"], [random variables ¶12 "Bayesian framework, model parameters, i.e., the weights and biases, are defined as random variables"]
calculate an estimate indicating a degree of being in the specific state based on a distribution of a plurality of the output values output from the data analysis model for the plurality of times. [calculates confidence and matrices based on a distribution and outputs (degree of being in a state (the prediction)) ¶5 " approximate the mean and covariance of each respective tensor normal distribution passing through the non-linear activation function of each non-linear perceptron"…" compute a confidence in the prediction based at least in part on the mean matrix and the covariance matrix of the output vector."]
Fathallah-Shaykh does not specifically teach input the input data to the data analysis model repeatedly a plurality of times;
However, Simons teaches input the input data to the data analysis model repeatedly a plurality of times; and [continual monitoring and evaluating ¶29, ¶64, deploys models regularly ¶17 "model parameters of the fault classification model trained in this way can then be transmitted to the motor vehicle once or at regular intervals." ]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the classification modeling disclosed by Fathallah-Shaykh by incorporating the input the input data to the data analysis model repeatedly a plurality of times disclosed by Simons because both techniques address the same field of machine learning and by incorporating Simons into Fathallah-Shaykh enables more statistically significant predictions with models[Simons ¶31, ¶62]
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh and Simons, as applied in the rejection of claim 1 above, and further in view of Adams et al. (US 20140358831 A1 hereinafter Adams)
As to dependent claim 2, Fathallah-Shaykh and Simons, teach the rejection of claim 1 that is incorporated.
Fathallah-Shaykh and Simons do not specifically teach wherein the circuitry is configured to calculate the estimate in such a manner that the degree of being in the specific state becomes lower as the distribution varies.
However, Adams teaches wherein the circuitry is configured to calculate the estimate in such a manner that the degree of being in the specific state becomes lower as the distribution varies. [high confidence becomes low confidence based on high variance ¶84 "Different amounts of uncertainty may be associated with estimates of machine learning system performance corresponding to different hyper-parameter values. For some hyper-parameter values the probabilistic model may be able to provide a high-confidence estimate (e.g., an estimate associated with a low variance) of the machine learning system's performance when configured with those hyper-parameter values, whereas for other hyper-parameter values the probabilistic model may provide a low-confidence estimate (e.g., an estimate associated with a high variance) of the machine learning system's performance when configured with those hyper-parameter values."]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh and Simons by incorporating the wherein the circuitry is configured to calculate the estimate in such a manner that the degree of being in the specific state becomes lower as the distribution varies disclosed by Adams because all techniques address the same field of machine learning systems and by incorporating Adams into Fathallah-Shaykh and Simons optimizes performance of models with more valid outputs [Adams ¶181]
Claims 3-6 are rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh, Simons and Adams, as applied in the rejection of claim 2 above, and further in view of Nagel et al. (US 10879831 B1 hereinafter Nagel)
As to dependent claim 3, Fathallah-Shaykh, Simons and Adams, teach the rejection of claim 2 that is incorporated.
Fathallah-Shaykh, Simons and Adams do not specifically teach wherein the circuitry is configured to: acquire control data of a motor control apparatus for a motor, the motor being configured to drive a mechanism of an industrial machine, input, as the input data, the control data to the data analysis model repeatedly a plurality of times, and calculate, as the specific state, an estimate indicating a degree of occurrence of an abnormal phenomenon specific.
However, Nagel teaches acquire control data of a motor control apparatus for a motor, the motor being configured to drive a mechanism of an industrial machine, [receives data (signals/signature) of a motor drive Col. 10 ln. 44-51 "the motor drive 30 may receive a control signal 116 from the industrial controller 10 such that different real-time signals 112 may be monitored at different points in a control program executing on the industrial controller 10. There may be, for example, periods of operation by the motor 50 controller by the motor drive during which a high torque is required for normal operation"]
input, as the input data, the control data to the data analysis model repeatedly a plurality of times, and [fed into an anomaly detector (model) Col. 12 ln. 44-67 "anomaly detector module 150 next compares the real-time signature generated to a set of expected signatures"]
calculate, as the specific state, an estimate indicating a degree of occurrence of an abnormal phenomenon specific. [calculates an estimate (percentage) Col. 12 ln. 44-67 "pattern matching routine may return, for example, a percentage likelihood that a real-time signature is a match"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh, Simons and Adams by incorporating the wherein the circuitry is configured to: acquire control data of a motor control apparatus for a motor, the motor being configured to drive a mechanism of an industrial machine, input, as the input data, the control data to the data analysis model repeatedly a plurality of times, and calculate, as the specific state, an estimate indicating a degree of occurrence of an abnormal phenomenon specific disclosed by Nagel because all techniques address the same field of machine learning systems and by incorporating Nagel into Fathallah-Shaykh, Simons and Adams improves detection of anomalies for a more real-time detection and inspection [Nagel Col. 2 ln 40-57]
As to dependent claim 4, the rejection of claim 3 is incorporated, Fathallah-Shaykh, Simons, Adams and Nagel further teach the memory is configured to store a plurality of the data analysis models configured to output, respectively, a plurality of the output values indicating whether a plurality of the abnormal phenomena that are mutually different occur, and [Simonis plurality of models that output different fault types (different abnormalities) ¶7 "a plurality of diagnostic models for a plurality of fault types, each of which is designed to detect a specific fault type in one of the components based on at least some of the plurality of operating parameters and to signal appropriate fault information associated with the fault type"]
the circuitry is configured to calculate an estimate indicating a degree of occurrence of each of the plurality of the abnormal phenomena that are mutually different based on a plurality of the distributions of, respectively, a plurality of output values, the plurality of output values being obtained by the input unit inputting the control data to each of the plurality of the data analysis models repeatedly a plurality of times. [Simonis provides appropriate degrees of anomalies ¶18-20 "provide an appropriate degree of anomaly as fault information"]
As to dependent claim 5, the rejection of claim 3 is incorporated, Fathallah-Shaykh, Simons, Adams and Nagel further teach identify a unit phenomenon that has occurred in the mechanism based on the control data, and [Simonis identifies fault types like a reconstruction error in an autoencoder ¶17-¶20]
input unit phenomenon data related to the unit phenomenon to the data analysis model repeatedly a plurality of times. [Simonis continual monitoring and evaluating ¶29, ¶64]
As to dependent claim 6, the rejection of claim 4 is incorporated, Fathallah-Shaykh, Simons, Adams and Nagel further teach identify a unit phenomenon that has occurred in the mechanism based on the control data, and [Simonis identifies fault types like a reconstruction error in an autoencoder ¶17-¶20]
input unit phenomenon data related to the unit phenomenon to the data analysis model repeatedly a plurality of times. [Simonis continual monitoring and evaluating ¶29, ¶64]
Claims 7, 10 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh, Simons and Adams, as applied in the rejection of claims 2 above, and further in view of Harper et al. (US 20210015417 A1 hereinafter Harper)
As to dependent claim 7, Fathallah-Shaykh, Simons and Adams, teach the rejection of claim 2 that is incorporated.
Fathallah-Shaykh, Simons and Adams do not specifically teach wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution.
However, Harper teaches wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh, Simons and Adams by incorporating the wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution disclosed by Harper because all techniques address the same field of machine learning systems and by incorporating Harper into Fathallah-Shaykh, Simons and Adams develop more efficient and effective ways of assisting humans [Harper ¶3-4]
As to dependent claim 10, the rejection of claim 7 is incorporated, Fathallah-Shaykh, Simons, Adams and Harper further teach input the training data to the data analysis model that is trained repeatedly a plurality of times; and [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
As to dependent claim 13, the rejection of claim 10 is incorporated, Fathallah-Shaykh, Simons, Adams and Harper further teach input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and [Harper confidence threshold at various levels ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
determine a largest candidate out of the candidates as the threshold. [Harper threshold set to make model less risky ¶91 "the model behaviour moves from risky to cautious but with less likelihood that a classification will be output (but with more certainty for classifications that are output). For binary classifications, there will always be at least 50% of the output distribution that will be within one of the two prediction zones, thus when α=0.5 the classification is determined by the median of the output distribution and a classification will always be made."]
Claims 8-9, 11-12 and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh, Simons, Adams and Nagel, as applied in the rejection of claims 3-4 above, and further in view of Harper
As to dependent claim 8, Fathallah-Shaykh, Simons, Adams and Nagel, teach the rejection of claim 3 that is incorporated.
Fathallah-Shaykh, Simons, Adams and Nagel do not specifically teach wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution.
However, Harper teaches wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh, Simons, Adams and Nagel by incorporating the wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution disclosed by Harper because all techniques address the same field of machine learning systems and by incorporating Harper into Fathallah-Shaykh, Simons, Adams and Nagel develop more efficient and effective ways of assisting humans [Harper ¶3-4]
As to dependent claim 9, Fathallah-Shaykh, Simons, Adams and Nagel, teach the rejection of claim 4 that is incorporated.
Fathallah-Shaykh, Simons, Adams and Nagel do not specifically teach wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution.
However, Harper teaches wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh, Simons, Adams and Nagel by incorporating the wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution disclosed by Harper because all techniques address the same field of machine learning systems and by incorporating Harper into Fathallah-Shaykh, Simons, Adams and Nagel develop more efficient and effective ways of assisting humans [Harper ¶3-4]
As to dependent claim 11, the rejection of claim 8 is incorporated, Fathallah-Shaykh, Simons, Adams, Nagel and Harper further teach input the training data to the data analysis model that is trained repeatedly a plurality of times; and [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
As to dependent claim 12, the rejection of claim 9 is incorporated, Fathallah-Shaykh, Simons, Adams, Nagel and Harper further teach input the training data to the data analysis model that is trained repeatedly a plurality of times; and [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
determine the threshold based on a distribution of a plurality of the output values output from the data analysis model that is trained for the plurality of times. [Harper confidence threshold at output distribution ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
As to dependent claim 14, the rejection of claim 11 is incorporated, Fathallah-Shaykh, Simons, Adams, Nagel and Harper further teach input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and [Harper confidence threshold at various levels ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
determine a largest candidate out of the candidates as the threshold. [Harper threshold set to make model risky ¶91 "the model behaviour moves from risky to cautious but with less likelihood that a classification will be output (but with more certainty for classifications that are output). For binary classifications, there will always be at least 50% of the output distribution that will be within one of the two prediction zones, thus when α=0.5 the classification is determined by the median of the output distribution and a classification will always be made."]
As to dependent claim 15, the rejection of claim 12 is incorporated, Fathallah-Shaykh, Simons, Adams, Nagel and Harper further teach input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and [Harper confidence threshold at various levels ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
determine a largest candidate out of the candidates as the threshold. [Harper threshold set to make model risky ¶91 "the model behaviour moves from risky to cautious but with less likelihood that a classification will be output (but with more certainty for classifications that are output). For binary classifications, there will always be at least 50% of the output distribution that will be within one of the two prediction zones, thus when α=0.5 the classification is determined by the median of the output distribution and a classification will always be made."]
Claims 16 is rejected under 35 U.S.C. 103 as being unpatentable over Fathallah-Shaykh and Simons, as applied in the rejection of claims 1 above, and further in view of Harper
As to dependent claim 16, the rejection of claim 1 is incorporated, Fathallah-Shaykh and Simons further teach input a plurality of pieces of the training data, respectively, to the data analysis model that is trained repeatedly a plurality of times, [Simonis trained and sent at regular intervals ¶21 "The model parameters of the anomaly detection model trained in this way can then be transmitted to the motor vehicle once or at regular intervals."]
Fathallah-Shaykh and Simons do not specifically teach calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data and determine a largest candidate out of the candidates as the threshold
However, Harper teaches calculate candidates for the threshold for each of the plurality of pieces of the training data based on the distribution obtained from the plurality of pieces of the training data, and [Harper confidence threshold at various levels ¶91 "Next a confidence threshold parameter a is used to tune predictions to a specified level of model uncertainty. For example, when α=0.95, at least 95% of the output distribution must lie in a given class zone in order for the input sample to be classified as belonging to that class (see FIG. 6). If this is not the case, then no prediction is made"]
determine a largest candidate out of the candidates as the threshold. [Harper threshold set to make model risky ¶91 "the model behaviour moves from risky to cautious but with less likelihood that a classification will be output (but with more certainty for classifications that are output). For binary classifications, there will always be at least 50% of the output distribution that will be within one of the two prediction zones, thus when α=0.5 the classification is determined by the median of the output distribution and a classification will always be made."]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the predictions disclosed by Fathallah-Shaykh and Simonis by incorporating the wherein the circuitry is configured to calculate the estimate based on a percentage of a plurality of the output values equal to or greater than a threshold or a plurality of the output values equal to or less than the threshold, in the distribution disclosed by Harper because all techniques address the same field of machine learning systems and by incorporating Harper into Fathallah-Shaykh and Simonis develop more efficient and effective ways of assisting humans [Harper ¶3-4]
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Applicant is required under 37 C.F.R. § 1.111(c) to consider these references fully when responding to this action.
Zafar et al. (US 20210209489 A1) teaches a classifier to detect anomalies and uses probability distributions (see ¶3)
It is noted that any citation to specific pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. A reference is relevant for all it contains and may be relied upon for all that it would have reasonably suggested to one having ordinary skill in the art. In re Heck, 699 F.2d 1331, 1332-33, 216 U.S.P.Q. 1038, 1039 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006, 1009, 158 U.S.P.Q. 275, 277 (C.C.P.A. 1968)).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Beau Spratt whose telephone number is 571 272 9919. The examiner can normally be reached 8:30am to 5:00pm (EST).
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, Jennifer Welch can be reached at 571 272 7212. The fax phone number for the organization where this application or proceeding is assigned is 571 483 7388.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866 217 9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800 786 9199 (IN USA OR CANADA) or 571 272 1000.
/BEAU D SPRATT/Primary Examiner, Art Unit 2143