DETAILED ACTION
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
2. Applicant's submission filed on 26 May 2026 [hereinafter Response] has been entered, where:
Claims 1-4 and 7-14, and 16-21 have been amended.
Claim 5 and 6 have been cancelled.
Claims 1-4 and 7-21 are pending.
Claims 1-4 and 7-21 are rejected.
Claim Rejections - 35 U.S.C. § 101
3. 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.
4. Claims 1-4 and 7-21 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to an abstract idea without significantly more.
Claim 1 recites a method, which is a process and thus one of the statutory categories of patentable subject matter. (35 U.S.C. § 101).
However, under Step 2A Prong One, the claim recites the limitations of “generating a pipeline graph having a plurality of layers including a feature scaling stage, a feature selection stage, and a classification stage,” “iteratively operating a plurality of pipelines through the pipeline graph over multiple rounds on a training dataset to determine a respective plurality of results,” “comparing the respective plurality of results to known results based on a predetermined metric,” and “identifying one or more leader pipelines based on the comparison.” The plain meaning of a “pipeline graph” is a visual representation of stages, steps, and flow of a process or workflow, showing how tasks, data, or materials move from one point to another. The activities of “generating a pipeline graph,” “iteratively operating,” “comparing,” and “identifying one or more leader pipelines,” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
The claim recites more details or specifics to the abstract idea of “generating a pipeline graph,” “wherein each layer of the plurality of layers has a plurality of machine learning components for performing a classification task,” and “the pipeline graph for the classification task comprises a specific number of a plurality of pipelines in the pipeline graph,” and accordingly, is merely more specific to the abstract idea. The plain meaning of “machine learning components” are modular, reusable, and self-contained pieces of code that perform a specific task in a pipeline graph simplify the development, testing, and deployment of ML workflows by breaking down complex tasks into smaller, manageable steps. Thus, “components for performing a predictive modeling task” can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, is a mental process. (MPEP § 2106.04(a)(2) sub III). The claim also recites more details or specifics to the abstract idea of “identifying one or more leader pipelines,” “wherein among the plurality of pipelines, results of the one or more leader pipelines are closest to the known results based on the parameter-level parallelism,” and “wherein the one or more leader pipelines are usable in an industrial application system for at least one of fault detection or anomaly prediction,” and accordingly, are merely more specific to the abstract idea.
Further, in relation to the activity of “iteratively operating,” the claim further recites limitations comprising “in a first round of the multiple rounds, operating, using a default hyperparameter, the plurality of machine learning components at a last layer of the pipeline graph on the training dataset,” and “in subsequent rounds of the multiple rounds: selecting progressively smaller portions of the plurality of machine learning components based on leading performance of a predictive model; and initiating hyperparameter tuning on each of the selected progressively smaller portions.” The activities of “operating” and “selecting” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
Also, the claim recites more details or specifics to the abstract idea of “iteratively operating,” wherein: “at least one pipeline is removed from the pipeline graph . . . based on the respective results of the previous round,” “a search space for identification of one or more leader pipelines is reduced based on the removal of the at least one pipeline of the plurality of pipelines within the pipeline graph,” “each pipeline of the plurality of pipelines indicates a distinct path,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space,” and “the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” and accordingly, are merely more specific to the abstract idea. Thus, claim 1 recites an abstract idea.
Under Step 2A Prong Two, the abstract idea of claim 1 is not integrated into a practical application, because the additional elements recited in the claim beyond the identified judicial exception include a “computer-implemented” method, where instructions to apply the abstract idea on generic computer components (i.e. the computer-implemented method) does not serve to integrate the abstract idea into a practical application. (MPEP § 2106.05(f)).
The claim also recites “[iteratively operating . . . wherein:] . . . an execution of each of the plurality of pipelines is accelerated by using a path-level parallelism or a parameter-level parallelism,” “the path-level parallelism includes executing the plurality of pipelines in parallel,” “the parameter-level parallelism includes executing each pipeline of the plurality of pipelines for different hyperparameter combinations in parallel,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space,” and “the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,”
which is the use of a generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application. Thus, claim 1 is directed to the abstract idea.
Finally, under Step 2B, the additional elements, taken alone or in combination, do not represent significantly more than the abstract idea itself. The additional elements recited in the claim include a “computer-implemented” method, where instructions to apply the abstract idea on generic computer components (i.e. the computer-implemented method) does not amount to significantly more than the abstract idea. (MPEP § 2106.05(f)). The claim also recites in relation to “[iteratively operating . . . wherein:] . . . an execution of each of the plurality of pipelines is accelerated by using a path-level parallelism or a parameter-level parallelism,” “the path-level parallelism includes executing the plurality of pipelines in parallel,” “the parameter-level parallelism includes executing each pipeline of the plurality of pipelines for different hyperparameter combinations in parallel,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space,” and “the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” which is the use of a generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not amount to significantly more than the abstract idea. Therefore, claim 1 is subject-matter ineligible.
Claim 14 recites a method, which is a process and thus one of the statutory categories of patentable subject matter. (35 U.S.C. § 101).
However, under Step 2A Prong One, the claim recites the limitations of “generating a pipeline graph having a plurality of layers including a feature scaling stage, a feature selection stage, and a classification stage,” “Iteratively operating each machine learning component of the plurality of machine learning components at a last layer of the plurality of layers over multiple rounds, on a training dataset,“ “removing, after each round, at least one machine learning component, of the plurality of machine-learning components, of the last layer that is not included in the selected portion for that round,” “adding an additional layer of the plurality of layers,” “identifying a plurality of extended pipeline paths using each machine learning component of the plurality of machine learning components of the additional one of the plurality of layers and each of the second portion,” and “selecting a third portion of the extended pipeline paths, the third portion being closest to the known results, wherein the third portion of the plurality of extended pipeline paths are usable in an industrial application system for at least one of fault detection or anomaly prediction.”
The plain meaning of a “pipeline graph” is a visual representation of stages, steps, and flow of a process or workflow, showing how tasks, data, or materials move from one point to another, which is subject to a mental process.
The activities of “generating,” “iteratively operating,” “selecting,” “removing,” “adding,” and “identifying” include limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, recite a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)).
Further, in relation to the activity of “iteratively operating,” the claim further recites limitations comprising “in a first round of the multiple rounds, operating, using a default hyperparameter, the plurality of machine learning components at the last layer of the pipeline graph on the training dataset” “selecting a first portion of the plurality of machine learning components of the last layer, the first portion being closest to a known result of the predictive modeling task,” “in a second round of the multiple rounds, initiating a first hyperparameter tuning on the first portion to determine a first tuned set of hyperparameters for the first portion of the plurality of machine learning components,” “selecting a second portion of the first portion, wherein the second portion being closest to the known result when the tuned set of hyperparameters are applied,” “model metrics of the second portion closest to the known results optimize a result,” “in a third round of the multiple rounds, initiating a second hyperparameter tuning on the second portion to determine a second tuned set of hyperparameters for the second portion,” “initiating a third hyperparameter tuning on the third power of the plurality of extended pipeline paths to determine a third tuned set of hyperparameters for each machine learning component.” The activities of “operating,” “selecting,” initiating,” and “model,” include limitations that that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, recite a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)).
The claim recites more details or specifics to the abstract idea of “removing . . . at least one machine learning component,” wherein “a search space for identification of a plurality of extended pipeline paths is reduced based on the removal of the at least one machine learning component of the plurality of machine learning components,” and accordingly, is merely more specific to the abstract idea.
The claim recites more details or specifics to the abstract idea of “generating a pipeline graph,” where “each layer of the plurality of layers has a plurality of machine learning components for performing a classification task,” and “the pipeline graph for the classification task comprises a specific number of a plurality of pipelines in the pipeline graph,” and accordingly, are merely more specific to the abstract idea. The plain meaning of “machine learning components” are modular, reusable, and self-contained pieces of code that perform a specific task in a pipeline graph simplify the development, testing, and deployment of ML workflows by breaking down complex tasks into smaller, manageable steps. Thus, “components for performing a predictive modeling task” can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, is a mental process. (MPEP § 2106.04(a)(2) sub III).
The claim also recites more details or specifics of the abstract ideas of “selecting a first portion,” where “the first portion being closest to a known result of the predictive modeling task,” “[selecting a second portion] . . . , the second portion being closest to the known result when the tuned set of hyperparameters are applied, and model metrics of the second portion closest to the known results optimize a result,” “[identifying the plurality of extended pipeline paths using each machine learning component of the plurality of machine learning components of the additional one of the plurality of layers and each of the second portion,” and “[selecting a third portion] . . . , the third portion being closest to the known result,” and “wherein the third portion of the plurality of extended pipeline paths are usable in an industrial application system for at least one of fault detection or anomaly prediction,” and accordingly, are merely more specific to the abstract idea.
Also, the claim recites more details or specifics to the abstract idea of “removing,” wherein “a search space for identification of a plurality of extended pipeline paths is reduced based on the removal of the at least one machine learning component of the one or more machine learning components,” and accordingly, is merely more specific to the abstract idea. Thus, claim 14 recites an abstract idea.
Under Step 2A Prong Two, the abstract idea of the claim is not integrated into a practical application, because the additional elements recited in the claim beyond the identified judicial exception include a “computer-implemented” method, where instructions to apply the abstract idea on generic computer components (i.e. the computer-implemented method) that does not integrate the abstract idea into a practical application. (MPEP § 2106.05(f)).
Further, the claim recites “initiating a first hyperparameter tuning on the first portion to determine a first tuned set of hyperparameters for the first portion of the one or more machine learning components,” “initiating a second hyperparameter tuning on the second portion to determine a second tuned set of hyperparameters for the second portion of the one or more machine learning components of the last layer;” “operating the plurality of extended pipeline paths on the training dataset with the default hyperparameters for each of the one or more machine learning components of the additional one of the plurality of layers,” and “initiating a third hyperparameter tuning on the third portion of the extended pipeline paths to determine a second tuned set of hyperparameters for each of the machine learning components of the additional one of the plurality of layers,” are activities directed to the use of generic computer components (computer-implemented method) to implement the abstract idea, which do not serve to integrate the abstract idea into a practical application. (MPEP § 2105(f)).
Also, the claim recites more details or specifics of the additional element of “initiating a third hyperparameter turning,” wherein “an execution of each of the plurality of extended pipeline paths is accelerated by using a path-level parallelism or a parameter-level parallelism,” “the path-level parallelism includes executing the plurality of extended pipeline paths in parallel,” “the parameter-level parallelism includes executing each of the plurality of extended pipeline paths for different hyperparameter combinations in parallel,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space, and the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” and accordingly, are merely more specific to the additional element. Thus, claim 14 is directed to the abstract idea.
Finally, under Step 2B, the additional elements, taken alone or in combination, do not represent significantly more than the abstract idea itself. The additional elements recited in the claim beyond the identified judicial exception include a “computer-implemented” method, where instructions to apply the abstract idea on generic computer components (i.e. the computer-implemented method) do not amount to significantly more than the abstract idea. (MPEP § 2106.05(f)).
Further, the claim recites “initiating a first hyperparameter tuning on the first portion to determine a first tuned set of hyperparameters for the first portion of the one or more machine learning components,” “initiating a second hyperparameter tuning on the second portion to determine a second tuned set of hyperparameters for the second portion of the one or more machine learning components of the last layer;” “operating the plurality of extended pipeline paths on the training dataset with the default hyperparameters for each of the one or more machine learning components of the additional one of the plurality of layers,” and “initiating a third hyperparameter tuning on the third portion of the extended pipeline paths to determine a second tuned set of hyperparameters for each of the machine learning components of the additional one of the plurality of layers,” are activities directed to the use of generic computer components (computer-implemented method) to implement the abstract idea, which do not amount to significantly more than the abstract idea. (MPEP § 2105(f)).
Also, the claim recites the additional element of “an execution of each of the plurality of extended pipeline paths is accelerated by using a path-level parallelism or a parameter-level parallelism,” which is the use of the generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application. The claim recites more details or specifics of the additional element of “an execution . . . is accelerated” where “the path-level parallelism includes executing the plurality of extended pipeline paths in parallel, the parameter-level parallelism includes executing each of the plurality of extended pipeline paths for different hyperparameter combinations in parallel, the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space, and the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” and accordingly, is merely more specific to the additional element. Therefore, claim 14 is subject-matter ineligible.
Claim 17 recites [a] non-transitory computer readable storage medium, which is product and thus one of the statutory categories of patentable subject matter. (35 U.S.C. § 101).
However, under Step 2A Prong One, the claim recites the limitations of “generating a pipeline graph having a plurality of layers including a feature scaling stage, a feature selection stage, and a classification stage,” “iteratively operating the plurality of pipelines through the pipeline graph over multiple rounds on a training dataset to determine a respective plurality of results,” “comparing the respective plurality of results to known results based on a predetermined metric,” and “identifying one or more leader pipelines based on the comparison.” The plain meaning of a “pipeline graph” is a visual representation of stages, steps, and flow of a process or workflow, showing how tasks, data, or materials move from one point to another. The activities of “generating a pipeline graph,” “iteratively operating,” “comparing,” and “identifying one or more leader pipelines,” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
The claim recites more details or specifics to the abstract idea of “generating a pipeline graph,” where “each layer of the plurality of layers having one or more machine learning components for performing a predictive modeling task,” and “the pipeline graph for the classification task comprises a specific number of a plurality of pipelines in the pipeline graph,” and accordingly, is merely more specific to the abstract idea. The plain meaning of “one or more machine learning components” are modular, reusable, and self-contained pieces of code that perform a specific task in a pipeline graph simplify the development, testing, and deployment of ML workflows by breaking down complex tasks into smaller, manageable steps. Thus, “components for performing a predictive modeling task” can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, is a mental process. (MPEP § 2106.04(a)(2) sub III).
The claim also recites more details or specifics to the abstract idea of “identifying one or more leader pipelines,” “wherein among the plurality of pipelines, results of the one or more leader pipelines are closest to the known results based on the parameter-level parallelism,” and “wherein the one or more leader pipelines are usable in in an industrial application system for at least one of fault detection or anomaly prediction,” and accordingly, are merely more specific to the abstract idea.
Further, in relation to the activity of “iteratively operating,” the claim further recites comprising “in a first round of the multiple rounds, operating, using a default hyperparameter, the plurality of machine learning components at a last layer of the pipeline graph on the training dataset,” “in subsequent rounds of the multiple rounds: selecting progressively smaller portions of the plurality of machine learning components based on leading performance of a predictive model: and initiating hyperparameter tuning on each of the selected progressively smaller portions,” and accordingly, are merely more specific to the abstract idea.
Also, the claim recited more details or specifics of the abstract idea of “iteratively operating,” wherein “at least one pipeline is removed from the pipeline graph . . . based on the respective results of the previous round,” “a search space for identification of one or more leader pipelines is reduced based on the removal of the at least one pipeline of the plurality of pipelines within the pipeline graph,” and accordingly, is merely more specific to the abstract idea. Thus, claim 17 recites an abstract idea.
Under Step 2A Prong Two, the abstract idea of claim 1 is not integrated into a practical application, because the additional elements recited in the claim beyond the identified judicial exception include a “non-transitory computer readable storage medium tangibly embodying a computer readable program code having computer readable instructions that , when executed, causes a computer device to carry out a method in providing computing efficiency of a computing device operating a pipeline execution engine”, where instructions to apply the abstract idea on generic computer components (i.e. the non-transitory computer readable storage medium, a computer device, a pipeline execution engine) do not serve to integrate the abstract idea into a practical application. (MPEP § 2106.05(f)).
The claim also recites “[iteratively operating . . . wherein:] . . . an execution of each of the plurality of pipelines is accelerated by using a path-level parallelism or a parameter-level parallelism,” which is the use of a generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application. The claim recites more details or specifics to the additional element of “an execution . . . accelerated by using,” where “the path-level parallelism includes executing the plurality of extended pipeline paths in parallel,” “the parameter-level parallelism includes executing each of the plurality of extended pipeline paths for different hyperparameter combinations in parallel,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space,” and “the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” and accordingly, is merely more specific to the additional element. Thus, claim 17 is directed to the abstract idea.
Finally, under Step 2B, the additional elements, taken alone or in combination, do not represent significantly more than the abstract idea itself. The additional elements recited in the claim include a “non-transitory computer readable storage medium tangibly embodying a computer readable program code having computer readable instructions that , when executed, causes a computer device to carry out a method in providing computing efficiency of a computing device operating a pipeline execution engine”, where instructions to apply the abstract idea on generic computer components (i.e. the non-transitory computer readable storage medium, a computer device, a pipeline execution engine) do not amount to significantly more than the abstract idea. (MPEP § 2106.05(f)).
The claim also recites “[iteratively operating . . . wherein:] . . . an execution of each of the plurality of pipelines is accelerated by using a path-level parallelism or a parameter-level parallelism,” which is the use of a generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not amount to significantly more than the abstract idea. The claim recites more details or specifics to the additional element of “an execution . . . accelerated by using,” where “the path-level parallelism includes executing the plurality of extended pipeline paths in parallel,” “the parameter-level parallelism includes executing each of the plurality of extended pipeline paths for different hyperparameter combinations in parallel,” “the different hyperparameter combinations correspond to a different hyperparameter point within a hyperparameter search space,” and “the parameter-level parallelism reduces a size of the hyperparameter search space after each round of the multiple rounds,” and accordingly, is merely more specific to the additional element. Therefore, claim 17 is subject-matter ineligible.
Claims 2 and 3 depend from claim 1, respectively. The claims recite more details or specifics of the abstract idea of the “generating a pipeline graph.” (claim 2: wherein the pipeline graph is generated from one or more default pipeline graphs for the classification task;” claim 3: “wherein: the plurality of machine learning components includes a no-operation component; and the training dataset passes without operation in a case where the pipeline includes the no-operation component”), and accordingly, are merely more specific to the abstract idea.
The additional elements of the claim does not serve to integrate the abstract idea into integrated into a practical application, (see MPEP § 2106.04(d)), nor do the additional elements amount to significantly more than the abstract idea, (MPEP § 2106.05 sub I; see also MPEP § 2106.05(a) – (h)), and thus, the claim recites no more than the abstract idea. Therefore, claims 2 and 3 are subject-matter ineligible.
Claim 4 depends from claim 1. Claim 18 depends from claim 17. The claims further recite an additional element of using the one or more machine learning components of the pipeline graph by applying a set of hyperparameters, (claims 4 and 18: “applying a set of hyperparameters to one or more of the selected ones of the plurality of machine learning components at each of the plurality of layers”), which is the use of the generic computer component (computer-implemented, non-transitory computer readable storage medium, computer device) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application, nor amounts to significantly more than the abstract idea. Therefore, claims 4 and 18 are subject-matter ineligible.
Claims 7 and 8 depend directly or indirectly from claim 1. The claims further recite the limitation of “selecting a first portion of the plurality of machine learning components of the last layer, the first portion a selection of the plurality of machine learning components of the last layer providing the leading performance of the predictive model.” The activity of “selecting a first portion” is a limitation that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)). Claim 8 recites more details or specifics of the abstract idea of “selecting a first portion” “wherein the first portion is about one-half of the plurality of machine learning components of the last layer,” and accordingly, is merely more specific to the abstract idea. The additional elements of the claim does not serve to integrate the abstract idea into integrated into a practical application, (see MPEP § 2106.04(d)), nor do the additional elements amount to significantly more than the abstract idea, (MPEP § 2106.05 sub I; see also MPEP § 2106.05(a) – (h)), and thus, the claim recites no more than the abstract idea. Therefore, claims 7 and 8 are subject-matter ineligible.
Claim 9 depends directly or indirectly from claim 1. The claim further recites “recites the additional element of “initiating a first hyperparameter tuning on the first portion to determine a first tuned set of hyperparameters for each of the first portion of the plurality of machine learning components,” which is a use of a generic computer component (computer-implemented method) to implement the abstract idea that does not integrate the abstract idea into a practical application, nor amount to significantly more than the abstract idea. (MPEP § 2106.05(f)). The claim also recites the limitation of “selecting a second portion of the first portion, the second portion providing leading performance of the predictive model,” which is a mental process, (MPEP § 2106.04(a)(2) sub III), and is one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)). Therefore, claim 9 is subject-matter ineligible.
Claim 10 depends directly or indirectly from claim 1. The claim recites more details or specifics of the abstract idea of “selecting,” “wherein the second portion is at least one-half of the plurality of the machine learning components of the first portion,” and accordingly, is merely more specific to the abstract idea. The additional elements of the claim does not serve to integrate the abstract idea into integrated into a practical application, (see MPEP § 2106.04(d)), nor do the additional elements amount to significantly more than the abstract idea, (MPEP § 2106.05 sub I; see also MPEP § 2106.05(a) – (h)), and thus, the claim recites no more than the abstract idea. Therefore, claim 10 is subject matter ineligible.
Claims 11 and 12 depend directly or indirectly from claim 1. The claims further recite “initiating a second hyperparameter tuning on the second portion to determine a second tuned set of hyperparameters for each of the second portion of the plurality of machine learning components of the last layer of the pipeline graph,” which is the use of generic computer components (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application, nor amounts to significantly more than the abstract idea. Claim 12 recites more details or specifics to the additional element of “initiating a first hyperparameter,” and “initiating a second hyperparameter” “wherein the first hyperparameter tuning and the second hyperparameter tuning both use a random search based hyperparameter tuning,” and accordingly, is merely more specific to the additional elements. The additional elements of the claim does not serve to integrate the abstract idea into integrated into a practical application, (see MPEP § 2106.04(d)), nor do the additional elements amount to significantly more than the abstract idea, (MPEP § 2106.05 sub I; see also MPEP § 2106.05(a) – (h)), and thus, the claim recites no more than the abstract idea. Thus, claims 11 and 12 are subject-matter ineligible.
Claim 13 depends directly or indirectly from claim 1. The claim recites further limitations of “adding an additional layer of the plurality of layers,” “identifying a plurality of extended pipeline paths using each machine learning component of the plurality of machine learning components of the additional layer of the plurality of layers and the second portion,” and “selecting a first portion of the extended pipeline paths, the first portion of the plurality of extended pipeline paths providing leading performance for the predictive model”. These activities of “adding,” “identifying,” and “selecting” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, are a mental process. (MPEP § 2106.04(a)(2) sub III). The claim also recites “operating the plurality of extended pipeline paths on the training dataset with the default hyperparameters for each machine learning component of the plurality of machine learning components of each of the plurality of expanded pipeline paths, wherein the default hyperparameters include the default hyperparameter,” and “initiating a third hyperparameter tuning on the first portion of the plurality of extended pipeline paths to determine a third tuned set of hyperparameters for each machine learning component of the plurality of machine learning components of the first portion of the plurality of extended pipeline paths,” which is the use of the generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application, nor amounts to significantly more than the abstract idea. Therefore, claim 13 is subject-matter ineligible.
Claims 15 and 16 depend from claim 14. Claim 15 recites more details or specifics of the additional element “initiating a first hyperparameter tuning,” and “initiating a second hyperparameter turning,” “wherein the first hyperparameter tuning and the second hyperparameter tuning both use a random search based hyperparameter tuning,” and accordingly, is merely more specific to the additional element. Claim 16 recites more details or specifics to the abstract idea of “selecting a first portion,” “wherein the first portion is about one-half of the plurality of machine learning components of the last layer, and accordingly, is merely more specific to the abstract idea. The abstract idea of the claim is not integrated into a practical application, (see MPEP § 2106.04(d)), nor does the claim amount to significantly more than the abstract idea, (MPEP § 2106.05), because the claim recites no more than the abstract idea. Therefore, claims 15 and 16 are subject-matter ineligible.
Claim 19 depends from claim 17. The claim further recites “selecting a first portion of the one or more machine learning components of the last layer, the first portion being closest to a known result of the predictive modeling task,” and “selecting a second portion of the first portion, the second portion being closest to the known result when the tuned set of hyperparameters are applied.” The activities of “selecting” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
The claim also recites additional elements of “operating each of the one or more machine learning components at a last layer of the pipeline graph on a training dataset using a default hyperparameter for each of the one or more machine learning components of the last layer,” “initiating a first hyperparameter tuning on the first portion to determine a tuned set of hyperparameters for each of the first portion of the one or more machine learning components,” and “initiating a second hyperparameter tuning on the second portion to determine a second tuned set of hyperparameters for each of the second portion of the one or more machine learning components of the last layer of the pipeline graph.” These limitations used the generic computer component (computer-implemented method) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application, nor does it amount to significantly more than the abstract idea. Thus, claim 19 is subject-matter ineligible.
Claim 20 depends directly or indirectly from claim 17. The claim recites further limitations of “adding an additional layer of the plurality of layers,” “identifying a plurality of extended pipeline paths using each machine learning component of the plurality of machine learning components of the additional layer of the plurality of layers and the second portion,” and “selecting a third portion of the extended pipeline paths, the third portion being closest to the known result.” These activities of “adding,” “identifying,” and “selecting” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
The claim also recites “operating the plurality of extended pipeline paths on the training dataset with the default hyperparameters for each of the plurality of machine learning components of the additional one of the plurality of layers, , wherein the default hyperparameters include the default hyperparameter” and “initiating a third hyperparameter tuning on the third portion of the extended pipeline paths to determine a second tuned set of hyperparameters for each machine learning component of the plurality of machine learning components of the additional one of the plurality of layers.” Using the generic computer component (computer-implemented method) to implement the abstract idea, (MPEP §2106.05(f)), that does not serve to integrate the abstract idea into a practical application, nor does it amount to significantly more than the abstract idea. (MPEP § 2106.05(f)). Thus, claim 20 is subject-matter ineligible.
Claim 21 depends from claim 1. The claim recites more details or specifics to the abstract idea of “comparing . . . based on a predetermined metric,” “wherein the predetermined metric corresponds to at least one of a time constraint, a memory constraint, an accuracy, or a precision of the predictive model,” and accordingly, is merely more specific to the abstract idea. Therefore, claim 21 is subject-matter ineligible.
Response to Arguments
5. Examiner has fully considered Applicant’s arguments, and responds below accordingly:
Claim Rejections – 35 U.S.C. § 101
6. Applicant submits “[t]he above-claimed features are inextricably tied to a computer technology. The pipeline graph for a classification task contains an extremely large number of possible machine-learning pipelines. As described in the Specification at paragraphs [0044] and [0056], a classification pipeline graph includes approximately 160,000 pipelines with 5 layers, 140+ nodes, and a hyperparameter grid with approximately 150 entries. This scale far exceeds the practical ability of a human to evaluate, compare, and optimize such combinations within a reasonable time.
Further, the claimed method requires iteratively operating through multiple rounds, first operating with default hyperparameters, then selecting progressively smaller portions based on leading performance, and initiating hyperparameter tuning on each selected portion. These operations involve actual execution of machine learning components on training data at a scale that cannot practically be performed in the human mind.
By exploring the pipeline graph by iteratively operating through multiple rounds for identifying optimal leader pipelines, the claimed method enables efficient discovery of highly accurate classification models. Further, the discovered optimal leader pipelines are for fault detection, thereby improving the reliability and speed of identifying faulty operating conditions, reducing downtime, preventing equipment failure, and improving overall operational safety and efficiency. Therefore, the claimed features cannot be considered as the alleged abstract idea of a Mental Process.” (Response at pp. 18-19).
Applicant also submits that: “Technical Problem: The extremely large combinatorial search space for a classification task (approximately 160,000 pipeline paths with approximately 150 hyperparameter entries) greatly exceeds the practical capability of the human mind. Technical Solution: An advantage of the claimed method is that the method provides a scalable mechanism for efficiently exploring the large pipeline graph using staged distributed optimization, iterative pruning, and hyperparameter tuning.” (Response at p. 21).
Also under Step 2A Prong Two, Applicant submits “Technical Problem: Fault detection and anomaly prediction in industrial application systems require optimized combinations of data transformations, feature-selection operations, and machine-learning models to accurately identify fault-related patterns. Technical Solution: An advantage of using the identified optimal leader pipelines for fault detection is that the leader pipelines provide optimized combinations of data transformations, feature-selection operations, machine-learning models, and tuned hyperparameters specifically suited for detecting fault-related patterns within industrial data.” (Response at p. 22).
Examiner Response:
Examiner respectfully submits that for Step 2A Prong One, the rejections identify the abstract idea (i.e., judicial exception) by referring to what is recited (i.e., set forth or described) in the claim and explain why it is considered an exception by identifying the abstract idea as it is recited and explain why it is an abstract idea. (MPEP § 2106.07(a)).
For example, the activities of “generating a pipeline graph,” “iteratively operating,” “comparing,” and “identifying one or more leader pipelines,” are limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP §2106.04(a)(2)).
Further, the plain meaning of a “pipeline graph” is a visual representation of stages, steps, and flow of a process or workflow, showing how tasks, data, or materials move from one point to another. Also, the plain meaning of “machine learning components” are modular, reusable, and self-contained pieces of code that perform a specific task in a pipeline graph simplify the development, testing, and deployment of ML workflows by breaking down complex tasks into smaller, manageable steps. Accordingly, these elements are within the grasp of the human mind.
Applicant submits that the “claimed features are inextricably tied to a computer technology.” (Response at p. 18). However, the claim is required to be directed to an improvement in the functionality of the computer, technology, or a technical field. That the operations of the pipeline graph of the claims could be performed more efficiently via a computer does not materially alter the patent eligibility of the claimed subject matter. (cf. Enfish, 822 F.3d at 1335-36 (distinguishing between claims wherein the focus of the claims is on an improvement in computer capabilities and those that invoke a computer as a tool)).
Also, no such technological advance or improvement to computer functionality is evident here. Rather, the claims merely employ generic computer components (computer-implemented method, computer-readable media, computer device) for computational efficiency in identifying leaders for a modeling task—i.e., the computer simply performs more efficiently what could otherwise be accomplished manually.
Accordingly, the rejections comply with the Office guidance for Step 2A Prong One, as set out above in detail.
Conclusion
7. 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.
8. The prior art made of record and not relied upon is considered pertinent to Applicant's disclosure:
(US Published Application 20200356726 to Nelson et al.) teaches that in addition to providing for pipeline generation, dependency graphs can be displayed and navigated by developers and business subject matter experts, which may make debugging and analysis easier. Further, the output from any natural language operation associated with any semantic or syntactic tags may be made available to any other the natural language operation and multiple natural language operations associated another semantic or semantic tags, by specifying it as a dependency. This allows for natural language understanding to be layered, such that low-level understanding is built up into higher and higher levels of understanding, where the higher levels depend on lower level understanding to be computed earlier in the natural language pipeline.
(Sugerman et al., “GRAMPS: A Programming Model for Graphics Pipelines,” ACM (2009)) teaches a programming model that generalizes concepts from modern real-time graphics pipelines by exposing a model of execution containing both fixed-function and application-programmable processing stages that exchange data via queues. GRAMPS allows the number, type, and connectivity of these processing stages to be defined by software, permitting arbitrary processing pipelines or even processing graphs.
9. Any inquiry concerning this communication or earlier communications from the Examiner should be directed to KEVIN L. SMITH whose telephone number is (571) 272-5964. Normally, the Examiner is available on Monday-Thursday 0730-1730.
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, KAKALI CHAKI can be reached on 571-272-3719. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
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.
/K.L.S./
Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights