Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
This Office action has been issued in response to amendment filed on 07/07/2026, Claims (1-10) and (11-20) are pending. Applicants' arguments have been carefully and respectfully considered and addressed. Accordingly, this action has been made FINAL necessitated by amendment.
Claims (1-10) and (11-20) are presented for examination.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 6, 11 and 16 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. The claims includes “first portion” or “second portion” which are not defined the in specification, it is unclear whether they pertain to the first or second variable or variable list.
Response to Arguments
Applicants' arguments have been carefully and respectfully considered and addressed. The arguments presented are moot based on amendment.
With regards to Applicant’s argument stating that da. Silva does not teach that optimizing the one or more associated variables included in the first variable list .... and it does not teach optimizing the one or more associated variables included in the first variable list locally at the one or more edge nodes of the federated learning simulation. Examiner respectfully disagrees, (Abstract, page. 2, paragraphs 2-4, page. 9, paragraphs 10-12, page. 10, paragraphs 1-2 wherein Xu describes parameters that are directed to edge nodes of the federated learning simulation, wherein the parameters include frozen parameters), (FIG. 2, [0020-0026], [0031] wherein da. Silva describes a local model optimizer that includes functionality to select local data pertinent to a given learning model and optimize it, wherein the local optimizer is part of the worker node as illustrated in FIG. 2).
Regarding Applicant’s argument stating that Xu does not teach the first variable list. Examiner respectfully disagrees. (FIG. 2, [0029], [0043] wherein da silva optimizes variables and parameters for nodes with edges as illustrated in FIG. 1), (FIG. 2, [0020-0026], [0031] wherein da. Silva describes a local model optimizer that includes functionality to select local data pertinent to a given learning model and optimize it, wherein the local optimizer is part of the worker node as illustrated in FIG. 2), ([0027] wherein da. Silva teaches collections of data, each representing tuples of feature-target data pertinent to optimizing a given learning model deployed on the worker node. Each feature-target tuple, of any given data collection, may refer to a finite ordered list (or sequence) of elements, including: a feature set; and one or more expected (target) classification or prediction values).
Regarding Applicant arguments that pertain to the amendment, the arguments were fully considered and are moot in view of the new ground rejection wherein da silva et al. US Patent Application Publication US 20220129786 A1 (hereinafter da silva) in view of in view of Xu, Ke et al. Foreign Patent Application Publication CN 114881229 A (hereinafter Xu) and further in view of Yu, Meng-Han et al. Foreign Application Publication CN 114648131 A (hereinafter Yu), wherein reference Yu teaches the amended limitations.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-6, 8-16 and 18-20 are rejected under AIA 35 U.S.C. 103(a) as being unpatentable over da silva et al. US Patent Application Publication US 20220129786 A1 (hereinafter da silva) in view of in view of Xu, Ke et al. Foreign Patent Application Publication CN 114881229 A (hereinafter Xu) and further in view of Yu, Meng-Han et al. Foreign Application Publication CN 114648131 A (hereinafter Yu).
Regarding claim 1, da silva teaches A method, comprising: for each federated learning simulation of a plurality of federated learning simulations (Abstract, [0001-0002], [0016-0020], [0042-0043] wherein da silva describes a system that includes a model simulation service that includes a plurality of first computer processors, wherein at least one first computer processor of the plurality of first computer processors is programmed to, for each federated learning simulation of a plurality of federated learning simulations, obtain a simulation configuration for the federated learning simulation, instantiate simulated nodes based on the simulation configuration, emulate learning interactions between the simulated nodes, probe the learning interactions to obtain a simulation analysis, and select, following a completion of the plurality of federated learning simulations, an optimal simulation configuration from a plurality of simulation configurations and based on a plurality of simulation analyses for the plurality of federated learning simulations) defining a machine learning model that is to be used in the federated learning simulation, the defined machine learning model having one or more associated variables, the defined machine learning model being implemented at one or more edge nodes of the federated learning simulation and at a central node of the federated learning simulation (FIG. 1, [0001], [0016-0021], [0042-0043] wherein da. Silva describes as illustrated in FIG. 1, defining a machine learning model for simulation that includes a central node and multiple nodes that represent worker nodes) defining a first variable list that specifies one or more of the associated variables that are to be optimized locally at the one or more edge nodes of the federated learning simulation (FIG. 2, [0029], [0043] wherein da silva optimizes variables and parameters for nodes with edges as illustrated in FIG. 1), (FIG. 2, [0020-0026], [0031] wherein da. Silva describes a local model optimizer that includes functionality to select local data pertinent to a given learning model and optimize it, wherein the local optimizer is part of the worker node as illustrated in FIG. 2) and aggregating at the central node of the federated learning simulation the one or more associated variables that are included in the second variable list and that are provided to the central node of the federated learning simulation by the one or more edge nodes of the federated learning simulation ([0018], [0020], [0033], [0036] wherein da. Silva incorporates a learning aggregator for aggregating the parameters at central node and performs a local optimization of a shared learning model using their respective local data. Thereafter, updates to the shared learning model, derived differently on each device based on different local data, may subsequently be forwarded to a federated learning coordinator (i.e., central node (104)), which aggregates and applies the updates to improve the shared learning model).
da. Silva teaches defining a second variable list that specifies one or more of the associated variables that are to be provided by the one or more edge nodes of the federated learning simulation to the central node of the federated learning simulation ([0018], [0029], [0046] wherein da silva describes an execution configuration that refer to a collection of parameters, which may be used to define an execution runtime of a given federation learning simulation. These parameters may include, but are not limited to, a seed used to initialize the behavior of (and thus shared between) simulated nodes of the given federated learning simulation, a number of optimization samples (i.e., feature-target tuples) to reside on each simulated worker node, a batch size or a subset number of optimization samples per batch for training and/or validation purposes, a proportion of optimization samples to be used during a training stage of learning model optimizations, a proportion of optimization samples to be used during a validation stage of learning model optimizations, a proportion of optimization samples on each simulated worker node to be sent to the simulated central node for testing purposes, and a learning rate of the simulated central node and/or simulated worker nodes. An execution configuration is not limited to the aforementioned parameter examples).
da. Silva teaches optimizing the one or more associated variables included in the first variable list locally (FIG. 2, [0020-0026], [0031] wherein da. Silva describes a local model optimizer that includes functionality to select local data pertinent to a given learning model and optimize it, wherein the local optimizer is part of the worker node as illustrated in FIG. 2).
da. Silva does not teach optimizing the one or more associated variables included in the first variable list locally at the one or more edge nodes of the federated learning simulation.
However in analogous art of federated learning method, Xu teaches optimizing the one or more associated variables included in the first variable list locally at the one or more edge nodes of the federated learning simulation (Abstract, page. 2, paragraphs 2-4, page. 9, paragraphs 10-12, page. 10, paragraphs 1-2 wherein Xu describes parameters that are directed to edge nodes of the federated learning simulation, wherein the parameters include frozen parameters).
It would have been obvious to a person in the ordinary skill in the art before the effective filing date of the claimed invention to combine Da. Silva with Xu by incorporating the method of optimizing the one or more associated variables included in the first variable list locally at the one or more edge nodes of the federated learning simulation of Xu into the method of defining a first variable list that specifies one or more of the associated variables that are to be optimized at the one or more edge nodes of the federated learning simulation of Da. Silva for the purpose of a personalized cooperative learning method and device based on parameter gradual freezing, wherein the method comprises: receiving the global model of this communication round time sent by the central server at the beginning of each communication round; splicing the global model and the local model of the previous communication round according to the mask matrix to obtain the local initial model of this communication round time; according to the variable parameter, determining the number of training rounds of this communication round, and training the local initial model of this communication round according to the number of training rounds, sending the local model of the communication round after finishing the training, (Xu: Abstract).
da. Silva does not teach wherein a first portion of the one or more associated variables included in the first variable list is not to be provided to the central node by the one or more edge nodes.
However in analogous art of federated learning method, Yu teaches wherein a first portion of the one or more associated variables included in the first variable list is not to be provided to the central node by the one or more edge nodes (Abstract, page. 2. ¶ 6-12, page. 4, ¶ 5-15, page. 6, ¶ 3-4, page. 9, ¶ 4, wherein Yu describes defining parameters of a local model of a participating edge node that can be as first potion and determines whether the local model reaches the accuracy requirement to determine whether to upload the model parameter to the central node.).
It would have been obvious to a person in the ordinary skill in the art before the effective filing date of the claimed invention to combine Da. Silva with Yu by incorporating the method of wherein a first portion of the one or more associated variables included in the first variable list is not to be provided to the central node by the one or more edge nodes of Yu into the method of defining a first variable list that specifies one or more of the associated variables that are to be optimized at the one or more edge nodes of the federated learning simulation of Da. Silva for the purpose of re-determining a plurality of participating edge nodes, sending a federal learning request to the plurality of participating edge nodes determined again, to make the newly determined plurality of participating edge nodes train new local model and determine new first participating edge node according to the new local model and model accuracy, receiving model data of new first participating edge node, and obtaining new global model according to model data of the new first participating edge node, until the model accuracy of the new global model reaches the preset accuracy requirement, (Yu: page. 3, ¶ 2).
Regarding claim 2, Da. Silva as modified by Xu and Yu teach for each federated learning simulation of the plurality of federated learning simulations: defining an aggregation map that includes an aggregation function that is used by the central node of the federated learning simulation to aggregate the one or more associated variables that are included in the second variable list and that are provided to the central node of the federated learning simulation by the one or more edge nodes of the federated learning simulation ([0042] wherein da silva describes a scenario configuration may refer to a collection of parameters, which may define a given federated learning simulation. These parameters may include, but are not limited to, a simulated central node class, a number of simulated worker nodes to employ, a behavior pattern for the simulated central node and/or worker nodes, an architecture (described below) for the learning model through which federated learning on the simulated central node and/or worker nodes may be performed, chosen datasets for emulating local data (described above) (see e.g., FIG. 2) on the simulated worker nodes, a chosen aggregation function employed by the simulated central node through which learning state from the various simulated worker nodes may be aggregated and used to update a global learning model, and a learning state compression technique to be employed by the simulated worker nodes when forwarding learning state to the simulated central node during the federated learning process. A scenario configuration is not limited to the aforementioned parameter examples).
Regarding claim 3, Da. Silva as modified by Xu and Yu teach wherein the one or more associated variables include one or more model variables that are part of the defined machine learning model and one or more variables that are related to, but are not directly part of, the defined machine learning model ([0042-009] wherein da silva describes a proportion of optimization samples to be used during a training stage of learning model optimizations, a proportion of optimization samples to be used during a validation stage of learning model optimizations, a proportion of optimization samples on each simulated worker node to be sent to the simulated central node for testing purposes, and a learning rate of the simulated central node and/or simulated worker nodes. An execution configuration is not limited to the aforementioned parameter examples).
Regarding claim 4, Da. Silva as modified by Xu and Yu teach wherein the one or more model variables include one or more of a weight variable, a bias variable, or a model statistical variable (claim 7 text, page. 4, paragraphs 2-6, page. 8, paragraphs 8-12, page. 9, paragraphs 1-3, page. 11, paragraph 9-10 wherein Xu teaches model variables that includes weight variable).
Regarding claim 5, Da. Silva as modified by Xu and Yu teach wherein the one or more related variables include one or more of statistical information, a number of samples used in model training, or information that is relevant to a particular edge node ([0042] wherein da silva teaches a number of simulated worker nodes to employ, a behavior pattern for the simulated central node and/or worker nodes, an architecture (described below) for the learning model through which federated learning on the simulated central node and/or worker nodes may be performed, chosen datasets for emulating local data (described above) (see e.g., FIG. 2) on the simulated worker nodes, a chosen aggregation function employed by the simulated central node through which learning state from the various simulated worker nodes may be aggregated and used to update a global learning model, and a learning state compression technique to be employed by the simulated worker nodes when forwarding learning state to the simulated central node during the federated learning process. A scenario configuration is not limited to the aforementioned parameter examples), (claims 1-2, 4, 10-13 text wherein Xu teaches variables for determining the number of training).
Regarding claim 6, Da. Silva as modified by Xu and Yu teach wherein a second portion of the one or more associated variables that are included in the first variable list includes standard variables that are optimized at the one or more edge nodes of the federated learning simulation and aggregated at the central node of the federated learning simulation (FIG. 2, [0029], [0043] wherein da silva optimizes variables and parameters for nodes with edges as illustrated in FIG. 1), ([0042] wherein da silva describes a scenario configuration may refer to a collection of parameters, which may define a given federated learning simulation. These parameters may include, but are not limited to, a simulated central node class, a number of simulated worker nodes to employ, a behavior pattern for the simulated central node and/or worker nodes, an architecture (described below) for the learning model through which federated learning on the simulated central node and/or worker nodes may be performed, chosen datasets for emulating local data (described above) (see e.g., FIG. 2) on the simulated worker nodes, a chosen aggregation function employed by the simulated central node through which learning state from the various simulated worker nodes may be aggregated and used to update a global learning model, and a learning state compression technique to be employed by the simulated worker nodes when forwarding learning state to the simulated central node during the federated learning process. A scenario configuration is not limited to the afromentioned parameter examples), (Claim 12 text, page. 8, ¶ 11, page. 9, ¶ 1, wherein Yu describes the federal learning request further comprises a model parameter to be trained of the global model, namely the initial parameter of the global model, to make each participating edge node training model parameter corresponding to the initial global model training model to obtain the local model of each participating edge node. exemplary, in the multi-wheel federation learning, the first round of federal learning training model parameter is initial value, the second wheel and the second wheel after the federal learning of the training model parameter is the model parameter of the global model of the previous round of federal learning, so as to continue to the federal learning on the basis of the previous round of federal learning, It improves the federal learning efficiency.
Regarding claim 8, Da. Silva as modified by Xu and Yu teach wherein the one or more associated variables that are only included in the second variable list are local information variables that are aggregated at the central node of the federated learning simulation, but are not optimized locally at the one or more edge nodes of the federated learning simulation (FIG. 2, [0029], [0043] wherein da silva optimizes variables and parameters for nodes with edges as illustrated in FIG. 1), ([0042] wherein da silva describes a scenario configuration may refer to a collection of parameters, which may define a given federated learning simulation. These parameters may include, but are not limited to, a simulated central node class, a number of simulated worker nodes to employ, a behavior pattern for the simulated central node and/or worker nodes, an architecture (described below) for the learning model through which federated learning on the simulated central node and/or worker nodes may be performed, chosen datasets for emulating local data (described above) (see e.g., FIG. 2) on the simulated worker nodes, a chosen aggregation function employed by the simulated central node through which learning state from the various simulated worker nodes may be aggregated and used to update a global learning model, and a learning state compression technique to be employed by the simulated worker nodes when forwarding learning state to the simulated central node during the federated learning process. A scenario configuration is not limited to the aforementioned parameter examples), (page. 6, ¶ 5-10, wherein Yu incorporates the test data can be set in the central node, so as to upload the model parameter on the edge node, using the test data to evaluate the accuracy of the edge model corresponding to the uploaded model parameter, and adjusting model parameter aggregate to the evaluation result).
Regarding claim 9, Da. Silva as modified by Xu and Yu teach wherein of the one or more associated variables include frozen variables that are not optimized locally at the one or more edge nodes of the federated learning simulation and are not aggregated at the central node of the federated learning simulation (Abstract, page. 2, paragraphs 2-4, page. 9, paragraphs 10-12, page. 10, paragraphs 1-2 wherein Xu describes parameters that are directed to edge nodes of the federated learning simulation, wherein the parameters include frozen parameters), ([0018], [0020], [0025-0026], [0029], [0034-0035] wherein da. Silva incorporates the local model optimizer for optimizing data locally and provides data to a learning state aggregator).
Regarding claim 10, Da. Silva as modified by Xu and Yu teach: selecting an optimal one of the federated learning simulations; and deploying the defined machine learning model, the central node, and the one or more edge nodes of the optimal one of the federated learning simulations on a plurality of computing systems (Abstract, [0015], [0020], [0040] wherein da silva facilitates communications between the model simulation service (400) and any other above-mentioned system component (e.g., worker nodes, central node, and/or client device (not shown)). To that extent, the service network interface (402) may include functionality to: enable an administrator to sign-in/login into the model simulation service (400) via the client device and, accordingly, allow the administrator to enter inputs and/or issue commands; relay the entered inputs and/or issued commands to the simulation configurator (404) and/or the simulation executor (406) for processing; obtain comparative visualizations from the simulation analyzer (408) for presentation to the administrator; select (either through artificial intelligence decisions or the administrator) an optimal federated learning configuration based, at least in part, on the comparative visualizations; and deploy at least a portion of the optimal federation learning configuration for real-world applications to the central and worker nodes. Further, one of ordinary skill will appreciate that the service network interface (402) may perform other functionalities without departing from the scope of the invention).
Regarding claim 11, da silva teaches A non-transitory storage medium having stored therein instructions that are executable by one or more hardware processors to perform operations comprising ([0017]). The claim is similar in scope to claim 1 therefore the claim is rejected under similar rationale.
Regarding claim 12, the claim is similar in scope to claim 2 therefore the claim is rejected under similar rationale.
Regarding claim 13, the claim is similar in scope to claim 3 therefore the claim is rejected under similar rationale.
Regarding claim 14, the claim is similar in scope to claim 4 therefore the claim is rejected under similar rationale.
Regarding claim 15, the claim is similar in scope to claim 5 therefore the claim is rejected under similar rationale.
Regarding claim 16, the claim is similar in scope to claim 6 therefore the claim is rejected under similar rationale.
Regarding claim 18, the claim is similar in scope to claim 8 therefore the claim is rejected under similar rationale.
Regarding claim 19, the claim is similar in scope to claim 9 therefore the claim is rejected under similar rationale.
Regarding claim 20, the claim is similar in scope to claim 10 therefore the claim is rejected under similar rationale.
Conclusion
THIS ACTION IS MADE FINAL. 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 extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HASSAN MRABI whose telephone number is (571)272-8875. The examiner can normally be reached on Monday-Friday, 7:30am-5pm. Alt, Friday, EST.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached on 571-270-5871. 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.
/HASSAN MRABI/Examiner, Art Unit 2144
Prosecution Insights