DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This action is responsive to the original application filed on 7/25/2024. Acknowledgment is made with respect to a claim of priority to Provisional Application 63/579,450 filed on 8/29/2023.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/process/file/efs/guidance/eTD-info-I.jsp.
Claim 1 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 1 of U.S. Patent Application No. 18/784,752. That is, claim 1 of U.S. Patent Application No. 18/784,752 discloses a method, processing a training data set of multiple class labels, training candidate classifiers with performance or training objectives, assessing the classifier to generate performance measurements, generating a tradeoff table, and providing a visualization of the tradeoff table.
Claim 3 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 2 of U.S. Patent Application No. 18/784,752. That is, claim 2 of U.S. Patent Application No. 18/784,752 training a set of candidate classifiers to maximize a Multidivergence objective.
Claim 8 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 8 of U.S. Patent Application No. 18/784,752. That is, claim 8 of U.S. Patent Application No. 18/784,752 discloses a computer program product, processing a training data set of multiple class labels, training candidate classifiers with performance or training objectives, assessing the classifier to generate performance measurements, generating a tradeoff table, and providing a visualization of the tradeoff table.
Claim 10 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 9 of U.S. Patent Application No. 18/784,752. That is, claim 9 of U.S. Patent Application No. 18/784,752 training a set of candidate classifiers to maximize a Multidivergence objective.
Claim 15 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 15 of U.S. Patent Application No. 18/784,752. That is, claim 15 of U.S. Patent Application No. 18/784,752 discloses a system, processing a training data set of multiple class labels, training candidate classifiers with performance or training objectives, assessing the classifier to generate performance measurements, generating a tradeoff table, and providing a visualization of the tradeoff table.
Claim 17 is provisionally rejected on the ground of nonstatutory double patenting as being anticipated by claim 16 of U.S. Patent Application No. 18/784,752. That is, claim 16 of U.S. Patent Application No. 18/784,752 training a set of candidate classifiers to maximize a Multidivergence objective.
This is a provisional double patenting rejection because the claims of the co-pending case (U.S. Patent Application No. 18/784,752) have not yet been patented.
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 8-14 are rejected under 35 U.S.C. § 101 because the claimed invention is directed towards non-statutory subject matter. The claims do not fall within at least one of the four categories of patent eligible subjected matter because the claimed invention is directed towards signals per se.
With respect to independent claim 8, the claim recites “a non-transient machine-readable medium.” Paragraph [0141] discloses:
“machine-readable medium” refers to any computer program product, apparatus and/or device, such as for example magnetic discs, optical disks, memory, and Programmable Logic Devices (PLDs), used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term “machine-readable signal” refers to any signal used to provide machine instructions and/or data to a programmable processor. The machine-readable medium can store such machine instructions non-transitorily, such as for example as would a non-transient solid-state memory or a magnetic hard drive or any equivalent storage medium. The machine-readable medium can alternatively or additionally store such machine instructions in a transient manner, such as for example as would a processor cache or other random access memory associated with one or more physical processor cores
However, the claim does not recite “non-transitory,” nor does the claim define the claimed “machine-readable medium” as excluding non-statutory propagating signals. Accordingly, claim 8 encompasses non-statutory subject matter. The specification [0141] does not equate “non-transient” to “non-transitory”. Rather, the specification distinguishes ‘non-transitorily’ from ‘non-transient’ and only uses a specific non-transient type, open-ended exemplary language as being ‘non-transitorily’. The BRI of non-transient still encompasses signal per se. Transient is generally defined as “lasting only a short time.” Transitory is “not permanent.” Therefore, under a broadest reasonable interpretation of the claim language, the “non-transient machine-readable medium” of claim 8 and its dependents may encompass transitory signals and is thus directed towards signals per se. Examiner suggests amending claim 8 and its dependents to recite "A computer program product comprising a non-transitory machine-readable medium". Dependent claims 9-14 depend on rejected claim 8 and are also rejected under 35 USC § 101 by virtue of this dependency. Appropriate correction is required.
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”).
When considering subject matter eligibility under 35 U.S.C. 101, it must be determined whether the claim is directed to one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter (Step 1). If the claim does fall within one of the statutory categories, the second step in the analysis is to determine whether the claim is directed to a judicial exception (Step 2A). The Step 2A analysis is broken into two prongs. In the first prong (Step 2A, Prong 1), it is determined whether or not the claims recite a judicial exception (e.g., mathematical concepts, mental processes, certain methods of organizing human activity). If it is determined in Step 2A, Prong 1 that the claims recite a judicial exception, the analysis proceeds to the second prong (Step 2A, Prong 2), where it is determined whether or not the claims integrate the judicial exception into a practical application. If it is determined at step 2A, Prong 2 that the claims do not integrate the judicial exception into a practical application, the analysis proceeds to determining whether the claim is a patent-eligible application of the exception (Step 2B). If an abstract idea is present in the claim, any element or combination of elements in the claim must be sufficient to ensure that the claim integrates the judicial exception into a practical application, or else amounts to significantly more than the abstract idea itself.
Claim 1
Step 1: The claim recites a method; therefore, it is directed to the statutory category of a process.
Step 2A Prong 1: The claim recites, inter alia:
processing, …, a training data set, wherein the training data set comprises a plurality of training examples, and wherein each of the plurality of training examples is associated with multiple class labels: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of broadly processing a training data set, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally process or evaluate data.
wherein random weights are assigned to multiple training objectives for training the set of candidate classifiers: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of assigning weights to training objectives, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally assign a weight of zero to an unwanted training objective.
assessing, …, each classifier of the set of candidate classifiers to generate multiple performance measurements for each of the set of candidate classifiers, wherein each of the multiple performance measurements is associated with each of the multiple training objectives, respectively: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of assessing classifiers to generate performance measurements, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally assess the output of a classifier and generate good or bad performance measurements for each classifier.
generating, by the at least one processor, a tradeoff table presenting performances of each classifier of the set of candidate classifiers, based at least in part on the multiple performance measurements: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of generating a tradeoff table or summary chart that presents classifier performances, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally generate a table or summary chart for classifiers using their respective performance measurements.
Step 2A Prong 2: The claim does not recite any additional limitations which integrate the abstract idea into a practical application. Specifically, the additional elements consist of “by at least one processor”, “training, by the at least one processor, a set of candidate classifiers”, “wherein the multiple training objectives corresponds to the multiple class labels”, “providing, by the at least one processor via a display, a visualization of the tradeoff table with dynamic interactive user experience to illustrate corresponding performances across different corresponding objectives of the multiple training objectives, wherein the dynamic interactive user experience allows a user to adjust a level of focus on a first set of the multiple training objectives”.
The additional elements of “by at least one processor” and “by the at least one processor” amount to generic computer components used as a tool to perform an existing process. The additional element “training, by the at least one processor, a set of candidate classifiers” amounts to reciting only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished because it is not clear how the generic candidate classifiers are broadly trained. Thus, the additional elements amount to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
The additional element providing, by the at least one processor via a display, a visualization of the tradeoff table with dynamic interactive user experience to illustrate corresponding performances across different corresponding objectives of the multiple training objectives, wherein the dynamic interactive user experience allows a user to adjust a level of focus on a first set of the multiple training objectives” is insignificant extra-solution activity required for any uses of the abstract ideas (see MPEP § 2106.05(g)).
The additional element of “wherein the multiple training objectives corresponds to the multiple class labels” amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use (see MPEP § 2106.05(h).
Thus, even when viewed individually and as an ordered combination, these additional elements do not integrate the abstract idea into a practical application and the claim is thus directed to the abstract idea.
Step 2B: Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea.
The additional elements of “by at least one processor” and “by the at least one processor” amount to generic computer components used as a tool to perform an existing process. The additional element “training, by the at least one processor, a set of candidate classifiers” amounts to reciting only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished because it is not clear how the generic candidate classifiers are broadly trained. Thus, the additional elements amount to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
The additional element “providing, by the at least one processor via a display, a visualization of the tradeoff table with dynamic interactive user experience to illustrate corresponding performances across different corresponding objectives of the multiple training objectives, wherein the dynamic interactive user experience allows a user to adjust a level of focus on a first set of the multiple training objectives” is insignificant extra-solution activity required for any uses of the abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Presenting offers and gathering statistics”; and see Cacioppi et al.; US 20140089234 A1; [0020]; “Some conventional systems describe interactive graphs for multi-objective functions that allow a user to interact with a visualization that describes the solutions to the multi-objective functions … Consequently, a system that enables a user to interact with a visualization to view the results of the optimization in real-time while the optimization of the multi-objective problem is still running may reduce the time needed to present worthwhile solutions to the user”).
The additional element of “wherein the multiple training objectives corresponds to the multiple class labels” amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use (see MPEP § 2106.05(h).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 2
Step 1: A process, as above.
Step 2A Prong 1: The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2, Step 2B: The additional element of “wherein the visualization of the tradeoff table dynamically presents a corresponding adjustment to a level of focus on a second set of the multiple training objectives in response to the user adjusting the level of focus on the first set of the multiple training objectives” amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use (see MPEP § 2106.05(h). Taken alone or in combination, the additional elements of the claim do not provide an inventive concept, integrate the abstract ideas into a practical application, or provide significantly more than the abstract ideas of the claim and thus the claim is subject-matter ineligible.
Claim 3
Step 1: A process, as above.
Step 2A Prong 1: The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2, Step 2B: The additional element of “training the set of candidate classifiers to maximize a randomized Multidivergence objective for each choice of the random weights assigned to the multiple training objectives, wherein the randomized Multidivergence objective is a weighted linear combination of divergence associated with the multiple training objectives” amounts to reciting only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished because it is not clear how the generic candidate classifiers are broadly trained to maximize an objective. Thus, the additional elements amount to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)). Taken alone or in combination, the additional elements of the claim do not provide an inventive concept, integrate the abstract ideas into a practical application, or provide significantly more than the abstract ideas of the claim and thus the claim is subject-matter ineligible.
Claim 4
Step 1: A process, as above.
Step 2A Prong 1: The claim recites, inter alia:
for a subset of the candidate classifiers, generating a set of mix-scores, wherein each of the mix-scores is a randomly weighted linear combination of customized scores generated by each of the subset of the candidate classifiers: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of generating mix scores for classifiers, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally generate a score for a classifier.
ranking the set of mix-scores to select a combination of weights associated with a highest ranked mix-score of the set of mix-scores: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of ranking scores, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper.
construct a preferred classifier using the selected combination of weights: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of generating a classifier using a weight, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally generate a model or classifier by assigning a weight of zero to unimportant information.
Step 2A Prong 2, Step 2B: The additional element of “wherein each of the subset of the candidate classifiers are trained to maximize a specific use-specific divergence related to a particular training objective of the multiple training objectives” amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use (see MPEP § 2106.05(h). Taken alone or in combination, the additional elements of the claim do not provide an inventive concept, integrate the abstract ideas into a practical application, or provide significantly more than the abstract ideas of the claim and thus the claim is subject-matter ineligible.
Claim 5
Step 1: A process, as above.
Step 2A Prong 1: The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2, Step 2B: The additional elements of “wherein the assessing each classifier of the set of candidate classifiers utilizes an assessment data set, and wherein the assessment data set comprises a plurality of evaluation examples, wherein each of the evaluation examples is associated with multiple evaluation class labels” amount to no more than generally linking the use of a judicial exception to a particular technological environment or field of use (see MPEP § 2106.05(h). Taken alone or in combination, the additional elements of the claim do not provide an inventive concept, integrate the abstract ideas into a practical application, or provide significantly more than the abstract ideas of the claim and thus the claim is subject-matter ineligible.
Claim 6
Step 1: A process, as above.
Step 2A Prong 1: The claim recites, inter alia:
adjusting influences of one or more features on outputs of the candidate classifiers by incorporating an additional objective into the multiple training objectives, wherein for the one or more features, the additional objective minimizes a distance between coefficients of the candidate classifiers and desired coefficients: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of adjusting influences of features on classifier outputs, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally adjust an influence of a feature by removing the feature.
Step 2A Prong 2, Step 2B: The claim does not recite any additional elements that are sufficient to integrate the judicial exceptions into a practical application or amount to significantly more than the judicial exception. As such, the claim is ineligible.
Claim 7
Step 1: A process, as above.
Step 2A Prong 1: The claim recites, inter alia:
incorporating a fairness objective into the multiple training objectives, wherein the fairness objective minimizes classifier outcome distribution disparities across different groups: Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of incorporating a fairness objective into other training objectives, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. For example, one can practically and mentally add more objectives to other objectives.
Step 2A Prong 2, Step 2B: The claim does not recite any additional elements that are sufficient to integrate the judicial exceptions into a practical application or amount to significantly more than the judicial exception. As such, the claim is ineligible.
Claims 8-14
Claims 8-14 recite a computer program product (step 1: would be a manufacture if the claims were directed towards a statutory category, but the claims are presently not directed towards a statutory category for the reasons discussed above) using a processor and non-transient machine-readable medium to perform the steps of claims 1-7, respectively, which by MPEP 2106.05(f) (“apply it”) cannot integrate an abstract idea into a practical application or provide significantly more than the abstract idea by itself, and are thus rejected for the same reasons set forth in the rejection of claims 1-7, respectively.
Claims 15-20
Claims 15-20 recite a system (step 1: a machine) using a processor and non-transient machine-readable medium to perform the steps of claims 1-6, respectively, which by MPEP 2106.05(f) (“apply it”) cannot integrate an abstract idea into a practical application or provide significantly more than the abstract idea by itself, and are thus rejected for the same reasons set forth in the rejection of claims 1-6, respectively.
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 (i.e., changing from AIA to pre-AIA ) 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, 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 1-3, 5, 8-10, 12, 15-17, and 19 are rejected under 35 USC § 103 as being obvious over Lin et al. (Lin et al., “Pareto Multi-Task Learning”, Dec. 30, 2019, arXiv:1912.12854v1, pp. 1-21, hereinafter “Lin”) in view of Wexler et al. (Wexler et al., “The What-If Tool: Interactive Probing of Machine Learning Models”, Jan. 2020, IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, VOL. 26, NO. 1, pp. 56-65, hereinafter “Wexler”).
Regarding claim 1, Lin discloses [a] computer-implemented method for generating a classifier, comprising: (Abstract; “we generalize this idea and propose a novel Pareto multi-task learning algorithm (Pareto MTL) to find a set of well-distributed Pareto solutions which can represent different trade-offs among different tasks”; and §1; and §6; note that the Experiments section is inherently implemented using a computer)
processing, by at least one processor, a training data set, wherein the training data set comprises a plurality of training examples, and wherein each of the plurality of training examples is associated with multiple class labels; (§6.1; “we randomly pick two images with different digits from the original MNIST dataset”, which discloses that the Pareto MTL classifier is applied to “MultiMNST”, in which the network must simultaneously predict both digit labels for every training example, and each training example carries multiple task or class labels)
training, by the at least one processor, a set of candidate classifiers, wherein random weights are assigned to multiple training objectives for training the set of candidate classifiers, and wherein the multiple training objectives corresponds to the multiple class labels; (Supplementary Material, §B; “We also compare Pareto MTL with the linear scalarization method with random weight research. As shown in sub-figure (b), Pareto MTL’s solutions dominate nearly all random search’s solutions”, which discloses generating multiple candidate models by scalarizing per-task losses with randomly generated weight vectors, with one candidate classifier resulting per random weight vector and the per-task losses corresponding to the class or task labels of the MultiMNST examples; and §4.2; and Algorithm 1; and Figure 14; and Supplementary Material, §F)
assessing, by the at least one processor, each classifier of the set of candidate classifiers to generate multiple performance measurements for each of the set of candidate classifiers, wherein each of the multiple performance measurements is associated with each of the multiple training objectives, respectively; (Supplementary Material, §B; “From sub-figure (a), it is clear that Pareto MTL with 25 preference vectors can represent the different trade-offs among three tasks much better. MTL practitioners can easily select their preferred solutions from the obtained results”, which discloses that each candidate model is evaluated and a performance value per task objective is produced, the performance values being per-task loss or accuracy values plotted for the trained candidate solutions; and §4.1; and Figures 4 and 13)
generating, by the at least one processor, a tradeoff table presenting performances of each classifier of the set of candidate classifiers, based at least in part on the multiple performance measurements; and (Supplementary Material, §B, Table 1; the table discloses generating a tradeoff table that presents performances or accuracies of the candidate classifiers (like Pareto MTL or MOO-MTL), and there are multiple performance measurements or accuracies accounted for; and Figures 4 and 6).
Lin fails to explicitly disclose but Wexler discloses providing, by the at least one processor via a display, a visualization [[of the tradeoff table]] with dynamic interactive user experience to illustrate corresponding performances across different corresponding objectives of the multiple training objectives, wherein the dynamic interactive user experience allows a user to adjust a level of focus on a first set of the multiple training objectives (Abstract; “The What-If Tool lets practitioners test performance in hypothetical situations, analyze the importance of different data features, and visualize model behavior across multiple models and subsets of input data. It also lets practitioners measure systems according to multiple ML fairness metrics”; and §4.3; “users can interact with configuration controls on the left side panel to specify ground truth features, select features to slice by, and set cost ratios for performance measures, and view the resulting performance table on the right side. An initial view of the performance table offers users performance measures tailored to the type of prediction task. Users can slice their data into subgroups by a single feature available in the dataset, or by the intersection of two features, in order to enable intersectional performance comparisons”, which discloses an interactive, on-display visualization in which a user manipulates a control associated with a subset of the plotted objective or “adjust a level of focus”, and the display or visualization responds to the user command).
Lin and Wexler are analogous art because both are concerned with evaluating the performance of multiple machine learning models. Before the effective filing date of the claimed invention, it would have been obvious to one skilled in machine learning to combine the visualization of Wexler with the training method and tradeoff tables of Lin to yield to the predictable result of providing, by the at least one processor via a display, a visualization of the tradeoff table with dynamic interactive user experience to illustrate corresponding performances across different corresponding objectives of the multiple training objectives, wherein the dynamic interactive user experience allows a user to adjust a level of focus on a first set of the multiple training objectives. The motivation for doing so would be to probe, visualize, and analyze ML systems (Wexler; Abstract).
Regarding claim 8, it is a computer program product claim corresponding to the steps of claim 1 and is rejected for the same reasons as claim 1.
Regarding claim 15, it is a system claim corresponding to the steps of claim 1 and is rejected for the same reasons as claim 1.
Regarding claims 2, 9, and 16, the rejection of claims 1, 8, and 15 are incorporated and Lin fails to explicitly disclose but Wexler discloses wherein the visualization of the tradeoff table dynamically presents a corresponding adjustment to a level of focus on a second set of the multiple training objectives in response to the user adjusting the level of focus on the first set of the multiple training objectives (§4.3.2; “In WIT, users can change the cost ratio, in order to automatically find the optimal positive classification threshold for a model given the results of model inference on the loaded dataset, in line with user needs”; and §4.3.3; “Classification thresholds for each slice can be individually modified so that the positive class threshold is different depending on the data point; the performance table and visualization instantly update to reflect resulting changes in performance. Applying a fairness optimization strategy through the tool automatically updates the classification thresholds for each slice individually in order to satisfy a particular fairness definition. Again, the results can be seen across the performance table with updates to thresholds, performance metrics and visualizations”).
The motivation to combine Lin and Wexler is the same as discussed above with respect to claim 1.
Regarding claims 3, 10, and 17, the rejection of claims 1, 8, and 15 are incorporated and Lin further discloses training the set of candidate classifiers to maximize a randomized Multidivergence objective for each choice of the random weights assigned to the multiple training objectives, wherein the randomized Multidivergence objective is a weighted linear combination of divergence associated with the multiple training objectives (§3.2, Equation 2; the equation discloses a scalarized MTL loss for weight vector wi, where the per-task loss L is the cross-entropy loss and wi is generated randomly; and see generally §3).
Regarding claims 5, 12, and 19, the rejection of claims 1, 8, and 15 are incorporated and Lin further discloses wherein the assessing each classifier of the set of candidate classifiers utilizes an assessment data set, and wherein the assessment data set comprises a plurality of evaluation examples, wherein each of the evaluation examples is associated with multiple evaluation class labels (Supplementary Material, §F; the section uses a test set to determine test accuracy, which is functionally equivalent to using an assessment data set that has evaluation examples associated with labels; and §6).
Claims 4, 11, and 18 are rejected under 35 USC § 103 as being obvious over Lin in view of Wexler and further in view of Caruana et al. (Caruana et al., “Ensemble selection from libraries of models”, Jul. 4, 2004, ICML '04: Proceedings of the twenty-first international conference on Machine learning, pp. 1-8, hereinafter “Caruana”).
Regarding claims 4, 11, and 18, the rejection of claims 1, 8, and 15 are incorporated and Lin further discloses wherein each of the subset of the candidate classifiers are trained to maximize a specific use-specific divergence related to a particular training objective of the multiple training objectives (§3, Equations 1-3).
Lin fails to explicitly disclose but Caruana discloses for a subset of the candidate classifiers, generating a set of mix-scores, wherein each of the mix-scores is a randomly weighted linear combination of customized scores generated by each of the subset of the candidate classifiers (§2.3; the section describes computing a combination score for a subset of models by weighting or averaging their output scores, the combination score is interpreted as the mix-scores)
ranking the set of mix-scores to select a combination of weights associated with a highest ranked mix-score of the set of mix-scores; and (§2.2-2.3; the hill-climbing/forward selection procedure computes a validation metric score for ensemble combinations and selects the combination whose score is best, which is functionally equivalent to ranking combinations and selecting a top-ranked combination)
construct a preferred classifier using the selected combination of weights (§1; “Return the ensemble from the nested set of ensembles that has maximum performance on the hillclimb(validation) set”).
Lin, Wexler, and Caruana are analogous art because all are concerned with evaluating the performance of multiple machine learning models. Before the effective filing date of the claimed invention, it would have been obvious to one skilled in machine learning to combine the ensembling of Caruana with the visualization of Wexler and the training method and tradeoff tables of Lin to yield to the predictable result of for a subset of the candidate classifiers, generating a set of mix-scores, wherein each of the mix-scores is a randomly weighted linear combination of customized scores generated by each of the subset of the candidate classifiers, wherein each of the subset of the candidate classifiers are trained to maximize a specific use-specific divergence related to a particular training objective of the multiple training objectives; ranking the set of mix-scores to select a combination of weights associated with a highest ranked mix-score of the set of mix-scores; and construct a preferred classifier using the selected combination of weights. The motivation for doing so would be to allow ensembles to be optimized to performance metric such as accuracy, cross entropy, mean precision, or ROC Area (Caruana; Abstract).
Claims 6, 13, and 20 are rejected under 35 USC § 103 as being obvious over Lin in view of Wexler and further in view of Li et al. (Li et al., “Explicit Inductive Bias for Transfer Learning with Convolutional Networks”, Jun. 6, 2018, arXiv:1802.01483v2, pp. 1-12, hereinafter “Li”).
Regarding claims 6, 13, and 20, the rejection of claims 1, 8, and 15 are incorporated and Lin fails to explicitly disclose but Li discloses adjusting influences of one or more features on outputs of the candidate classifiers by incorporating an additional objective into the multiple training objectives, wherein for the one or more features, the additional objective minimizes a distance between coefficients of the candidate classifiers and desired coefficients (§3, Equation 2; the equation discloses an SP penalty terms that is incorporated into an overall loss that constrains or adjusts the influence of input features or layers learned coefficients).
Lin, Wexler, and Li are analogous art because all are concerned with evaluating the performance of multiple machine learning models. Before the effective filing date of the claimed invention, it would have been obvious to one skilled in machine learning to combine the influence adjustments of Li with the visualization of Wexler and the training method and tradeoff tables of Lin to yield to the predictable result of adjusting influences of one or more features on outputs of the candidate classifiers by incorporating an additional objective into the multiple training objectives, wherein for the one or more features, the additional objective minimizes a distance between coefficients of the candidate classifiers and desired coefficients. The motivation for doing so would be ti retain features for a source task during transfer learning (Li; Abstract).
Claims 7 and 14 are rejected under 35 USC § 103 as being obvious over Lin in view of Wexler and further in view of Zafar et al. (Zafar et al., “Fairness Constraints: Mechanisms for Fair Classification”, Mar. 23, 2017, arXiv:1507.05259v5, pp. 1-12; hereinafter “Zafar”).
Regarding claims 7 and 14, the rejection of claims 1 and 8 are incorporated and Lin fails to explicitly disclose but Zafar discloses wherein training the set of candidate classifiers further comprises incorporating a fairness objective into the multiple training objectives, wherein the fairness objective minimizes classifier outcome distribution disparities across different groups (§3, “we first introduce our measure of decision boundary (un)fairness, the decision boundary covariance. We then derive two complementary formulations. The first formulation ensures compliance with disparate impact doctrine in its basic form (ensure a given p%-rule) by maximizing accuracy subject to fairness constraints. The second formulation guarantees fulfilling disparate impact’s “business necessity” clause by maximizing fairness subject to ac curacy constraints”, which discloses incorporating fairness objectives or constraints into training objectives that minimizes classifier outcome distribution disparities; and §3.2).
Lin, Wexler, and Zafar are analogous art because all are concerned with evaluating the performance of multiple machine learning models. Before the effective filing date of the claimed invention, it would have been obvious to one skilled in machine learning to combine the fairness objectives of Zafar with the visualization of Wexler and the training method and tradeoff tables of Lin to yield to the predictable result of wherein training the set of candidate classifiers further comprises incorporating a fairness objective into the multiple training objectives, wherein the fairness objective minimizes classifier outcome distribution disparities across different groups. The motivation for doing so would be to design fair classifiers by leveraging a novel intuitive measure of decision boundary (un)fairness (Zafar; Abstract).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Brent Hoover whose telephone number is (303)297-4403. The examiner can normally be reached Monday - Friday 9-5 MST.
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, Abdullah Kawsar can be reached at 571-270-3169. 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.
/BRENT JOHNSTON HOOVER/ Primary Examiner, Art Unit 2127
