DETAILED ACTION
Remarks
This office action is issued in response to communication filed on 10/18/23. Claims 1-11 are pending in this Office Action.
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 .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-11 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1, 4 and 11:
Step 1: Statutory Category ?: Yes. claims 1 and 4 recite a system (i.e., a “machine”) and claim 11 recites a method (i.e., a “process”) which are statutory categories.
Claim 1:
Step 2A-Prong 1: Judicial Exception Recited ?: Yes. Claim 1 recites one or more limitations that can be performed in human mind using observation, evaluation, judgment and opinion including with the help of a pen and paper.
“in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, and calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets”,( Bold text emphasis added. Except for the language of “similarity calculation model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
in the second calculation process, each of the plurality of client terminals calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model, and outputs a learning weight by performing a similarity calculation between the first similarity and the second similarity, ( Bold text emphasis added. Except for the language of “similarity calculation model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
in the first learning process, each of the plurality of client terminals learns a first analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set based on the first analysis model, the first explanatory variable and the first objective variable of the learning data set, and the learning weight ( Bold text emphasis added. Except for the language of “a first analysis model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
, and
transmits a first learning result to the server, and in the first integration process, the server generates a second analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set by integrating the first learning result by the first learning process from the plurality of client terminals. (Bold text emphasis added. Except for the language of “second analysis model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
Step 2A-Prong 2: Integrated into a practical application? No.
Claim 1 recites additional elements :
a plurality of client terminals that respectively have learning data sets, and a server capable of communicating with the plurality of client terminals, ( data gathering step and therefore are insignificant extra-solution activities. (See MPEP 2106.05(g)).
the federated learning system performing federated learning, in which each of the plurality of client terminals learns a model using each of the learning data sets and the server repeats a process of integrating the model of each of the plurality of client terminals using the learning result, wherein the learning data set includes one or more data samples including a client ID specifying the client terminal, a first explanatory variable, and a first objective variable, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
the federated learning system executes a first calculation process by the server and a second calculation process by each of the plurality of client terminals ( mere instructions to apply the exception using generic computer ), and
executes a first federated learning process that repeats a first learning process by each of the plurality of client terminals and a first integration process by the server until a first end condition is satisfied, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, (data gathering step and therefore are insignificant extra-solution activities. (See MPEP 2106.05(g)) and calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the second calculation process, each of the plurality of client terminals calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model, and outputs a learning weight by performing a similarity calculation between the first similarity and the second similarity, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the first learning process, each of the plurality of client terminals learns a first analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set based on the first analysis model, the first explanatory variable and the first objective variable of the learning data set, and the learning weight, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception) and
transmits a first learning result to the server ((data gathering step and therefore are insignificant extra-solution activities. (See MPEP 2106.05(g)), and in the first integration process, the server generates a second analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set by integrating the first learning result by the first learning process from the plurality of client terminals. ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
Step 2B: Recites additional elements that amount to significantly more than the judicial exception? No.
Claim 1 does not include additional elements that are sufficient to amount to significantly more than judicial exception. As indicates above, the additional element of data gathering is well-understood, routine conventional activities previously known to the industry and therefore do not amount to significantly more than the judicial exception. (See MPEP 2106.05(d)) and 2106.07(a)III). The client, server and models are at best the equivalent of adding the words “apply it “ to the exception. Even when consider in combination, the additional elements do not provide an inventive concept, claim 1 therefore ineligible.
Claim 2 recites additional element “ wherein each of the plurality of client terminals executes the first learning process using the second analysis model generated by the first integration process as the first analysis model” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 2 therefore ineligible.
Claim 3 recites additional element “ wherein the server executes inference by inputting the input data set into the second analysis model when the first end condition is satisfied” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 3 therefore ineligible.
Claim 4:
Step 2A-Prong 1: Judicial Exception Recited ?: Yes. Claim 1 recites one or more limitations that can be performed in human mind using observation, evaluation, judgment and opinion including with the help of a pen and paper.
the federated learning system executes a first calculation process by the server and a second calculation process by each of the plurality of client terminals, and executes a first federated learning process that repeats a first learning process by each of the plurality of client terminals and a first integration process by the server until a first end condition is satisfied, (Bold text emphasis added. Except for the language of “server and client”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, and calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets, (Bold text emphasis added. Except for the language of “similarity calculation model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
in the second calculation process, each of the plurality of client terminals calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model for calculating the similarity between the data sample and the plurality of learning data sets (Bold text emphasis added. Except for the language of “similarity calculation model”), there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process)
in the first learning process, each of the plurality of client terminals learns an individual analysis model for calculating a predicted value of the first objective variable from the first explanatory variable based on the individual analysis model, the first explanatory variable, the first objective variable, and a specific second similarity with a specific learning data set calculated by each of the plurality of client terminals by the calculation process (Bold text emphasis added. Except for the language of “analysis model and client terminals” , there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process), and
in the first integration process, the server generates an analysis model for calculating the predicted value of the first objective variable from the first explanatory variable for the input data set by integrating a plurality of first learning results by the first learning process from the plurality of client terminals, using the first similarity. (Bold text emphasis added. Except for the language of “analysis model and client terminals” , there is nothing in the claim that prevents the user from perform the calculation in his or her mind. This is a mental process),
Step 2A-Prong 2: Integrated into a practical application? No.
Claim 4 recites additional elements :
a plurality of client terminals that respectively have learning data sets, and a server capable of communicating with the plurality of client terminals, ( data gathering step and therefore are insignificant extra-solution activities. (See MPEP 2106.05(g)).
the federated learning system performing federated learning, in which each of the plurality of client terminals learns a model using each of the learning data sets and the server repeats a process of integrating the model of each of the plurality of client terminals using the learning result, wherein the learning data set includes one or more data samples including a client ID specifying the client terminal, a first explanatory variable, and a first objective variable ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
,the federated learning system executes a first calculation process by the server and a second calculation process by each of the plurality of client terminals, and executes a first federated learning process that repeats a first learning process by each of the plurality of client terminals and a first integration process by the server until a first end condition is satisfied ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, (data gathering step and therefore are insignificant extra-solution activities. (See MPEP 2106.05(g)) and
calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets, ( mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the second calculation process, each of the plurality of client terminals calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model for calculating the similarity between the data sample and the plurality of learning data sets (mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
in the first learning process, each of the plurality of client terminals learns an individual analysis model for calculating a predicted value of the first objective variable from the first explanatory variable based on the individual analysis model, the first explanatory variable, the first objective variable, and a specific second similarity with a specific learning data set calculated by each of the plurality of client terminals by the calculation process (mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception), and
in the first integration process, the server generates an analysis model for calculating the predicted value of the first objective variable from the first explanatory variable for the input data set by integrating a plurality of first learning results by the first learning process from the plurality of client terminals, using the first similarity. (mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception)
Step 2B: Recites additional elements that amount to significantly more than the judicial exception? No.
Claim 4 does not include additional elements that are sufficient to amount to significantly more than judicial exception. As indicates above, the additional element of data gathering is well-understood, routine conventional activities previously known to the industry and therefore do not amount to significantly more than the judicial exception. (See MPEP 2106.05(d)) and 2106.07(a)III). The client, server and models are at best the equivalent of adding the words “apply it “ to the exception. Even when consider in combination, the additional elements do not provide an inventive concept, claim 4 therefore ineligible.
Claim 5 recites additional element : “wherein each of the plurality of client terminals executes the first learning process using the analysis model generated by the first integration process as the individual analysis model” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 5 therefore ineligible.
Claim 6 recites additional element : “ wherein the server executes inference by inputting the input data set into the analysis model when the first end condition is satisfied” ” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 6 therefore ineligible.
Claim 7 recites additional element : “ wherein in the second calculation process, each of the plurality of client terminals calculates the second similarity by inputting a combination of the first explanatory variable and the first objective variable of the data sample into the similarity calculation model as a second explanatory variable” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 7 therefore ineligible.
Claim 8 recites additional element : “ wherein in the second calculation process, each of the plurality of client terminals calculates a learning weight according to the specific second similarity, and in the first learning process, each of the plurality of client terminals learns the individual analysis model based on the individual analysis model, the first explanatory variable, the first objective variable, and a learning weight corresponding to the specific second similarity calculated by each of the plurality of client terminals by the second calculation process” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 8 therefore ineligible.
Claim 9 recites additional element : “ wherein the federated learning system executes, before the first federated learning process, a second federated learning process in which a second learning process by each of the plurality of client terminals and a second integration process by the server are repeated until a second end condition is satisfied, in the second learning process, each of the plurality of client terminals learns a similarity calculation model to be learned using a combination of the first explanatory variable and the first objective variable as a second explanatory variable and the client ID as a second objective variable, and in the second integration process, the server integrates second learning results of the similarity calculation models to be learned from the plurality of client terminals by the second learning process and generates an integrated similarity calculation model obtained by integrating the similarity calculation models to be learned of the plurality of client terminals as the similarity calculation model” ” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 9 therefore ineligible.
Claim 10 recites additional element : “ wherein in the second learning process, each of the plurality of client terminals sets the updated similarity calculation models to be learned as the similarity calculation model when the second end condition is satisfied” ” which is mere instructions to apply the exception using generic computer or merely uses a computer as a tool to perform the exception and at best the equivalent of adding the words “apply it “ to the exception. . Even when consider in combination, the additional elements do not provide an inventive concept, claim 10 is therefore ineligible.
Claim 11 recites a method with similar features of the system of claim 1 and therefore rejected for the rationale as indicates in the above rejection of claim 1.
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 1-11 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.
Claim 1 recites the limitation " in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, and calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets". There is insufficient antecedent basis for this limitation in the claim because “plurality of learning data sets” has not been mentioned previously in the claim. Furthermore, it is not clear how the calculation is done if the server does not have “the plurality of learning data sets”. The server only receives/obtains an input data set, NOT “the plurality learning data sets”. Claim 1 only recites “ a plurality of client terminals that respectively have learning data sets, and a server capable of communicating with the plurality of client terminals”. There is no mention in the claim that the server comprise “a plurality learning data sets” or obtains a plurality learning data sets ( assuming from each of the client terminals). Similarly, the claim 1 further recites “in the second calculation process, each of the plurality of client terminals calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model”. It is not clear how the client terminal can calculates a second similarity using “the plurality of learning data sets” wherein the only data set the client has is the respective data sets (a plurality of client terminals that respectively have learning data sets). Appropriate correction is required.
Independent claims 4 and 11 recite similar limitations and therefore having the same issue.
Claims 2-3 and 5-10 are also indefinite due to its dependency to claims 1 or 4.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-11 are rejected under 35 U.S.C. 103 as being unpatentable over Uehara. (US Patent Application Publication 2022/0237898 A1, hereinafter “Uehara”) and further in view of Dimitriadis et al.(US Patent Application Publication 2022/0036178 A1, hereinafter “Dimitriadis”)
As to claim 1, Uehara teaches a federated learning system comprising a plurality of client terminals that respectively have learning data sets, and a server capable of communicating with the plurality of client terminals, (Uehara par [0012] teaches a machine learning system includes a plurality of client terminals and an integration server)
the federated learning system performing federated learning, in which each of the plurality of client terminals learns a model using each of the learning data sets ( Uehara par [0020] teaches each client holds local data LD in local storage) and the server repeats a process of integrating the model of each of the plurality of client terminals using the learning result (Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy) , wherein the learning data set includes one or more data samples including a client ID specifying the client terminal, a first explanatory variable, and a first objective variable, (Uehara Fig.3 and par [0064] teaches client ID, imaging condition, learning result )
the federated learning system executes a first calculation process by the server and a second calculation process by each of the plurality of client terminals , and executes a first federated learning process that repeats a first learning process by each of the plurality of client terminals and a first integration process by the server until a first end condition is satisfied,( Uehara par [0072] teaches local LD model . Uehara par [0076] teaches the integration server receives each learning result and the imaging condition information from the plurality of clients. Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy)
in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, and [ calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets] , (Uehara par [0084] teaches the master model learning management program of the integration server evaluates the inference accuracy of the master model candidate MMC using verification data for each of various imaging condition categories prepared in advance)
in the second calculation process, each of the plurality of client terminals [calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model, and outputs a learning weight [by performing a similarity calculation between the first similarity and the second similarity], (Uehara par [0081] teaches the data of the local model LM as a learning results provided from each client to the integration server may be the weight parameter of the trained local model LM. . Uehara par [0087] teaches the client CL1 performs the training of a local model LM11 using the local data LD11 as the learning data)
in the first learning process, each of the plurality of client terminals learns a first analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set based on the first analysis model, the first explanatory variable and the first objective variable of the learning data set, and the learning weight, and transmits a first learning result to the server (Uehara par [0081] teaches the data of the local model LM as a learning results provided from each client to the integration server may be the weight parameter of the trained local model LM. Uehara par [0092] teaches the learning results from clients are transmitted to the server) , and
in the first integration process, the server generates a second analysis model that calculates a predicted value of the first objective variable from the first explanatory variable for the input data set by integrating the first learning result by the first learning process from the plurality of client terminals. (Uehara par [0094] teaches the master model candidate creation unit integrates the learning results of the client clusters for each imaging condition category classified by the imaging condition classification unit to create a master model candidate MMck.)
Uehara fails to expressly teach calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets; calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model, and outputs a learning weight by performing a similarity calculation between the first similarity and the second similarity.
However, Dimitriadis teaches calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets; (Dimitriadis par [0019] the data sets 110, 112, and 114 may include labels or other indicators of classification of the data (e.g., ‘ground truths’ associated with the data set or other forms of data that may be compared to model outputs of the global model) that can be used in combination with the model outputs 105, 107, and 109 to determine the accuracy and/or associated loss functions of the global model 102 with respect to each data set)
calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model (Dimitriadis par [0019] the data sets 110, 112, and 114 may include labels or other indicators of classification of the data (e.g., ‘ground truths’ associated with the data set or other forms of data that may be compared to model outputs of the global model) that can be used in combination with the model outputs 105, 107, and 109 to determine the accuracy and/or associated loss functions of the global model 102 with respect to each data set), and outputs a learning weight by performing a similarity calculation between the first similarity and the second similarity. (Dimitriadis par [0043] teaches at least one gradient quality metric is determined for each gradient of the plurality of gradients. In some examples, the determined gradient quality metrics include mean values of the gradients, variance values of the gradients, and/or training loss values of the gradients. Par [0044] teaches [0044] At 408, a plurality of weight factors is calculated based on the gradient quality metrics. In some examples, the gradient quality metrics of each gradient are compared to the general set of gradient quality metrics of the plurality of gradients to calculate a weight factor for the gradient relative to the other gradients of the plurality of gradients)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teaching of Uehara and Dimitriadis to achieve the claimed invention . One would have been motivated to make such combination to increase accuracy or performance of the model .( Dimitriadis par [0029])
As to claim 2, Uehara and Dimitriadis teach the federated learning system according to claim 1, wherein each of the plurality of client terminals executes the first learning process using the second analysis model generated by the first integration process as the first analysis model. (Uehara par [0100] teaches the integration server synchronizes the latest version of the master model with the local model LM on each client before each clients starts the training)
As to claim 3, Uehara and Dimitriadis teach the federated learning system according to claim 1, wherein the server executes inference by inputting the input data set into the second analysis model when the first end condition is satisfied. (Uehara par [0105] teaches causing the master model to perform an inference by using the verification data as an input to the master model )
As to claim 4, Uehara teaches a federated learning system comprising a plurality of client terminals that respectively have learning data sets, and a server capable of communicating with the plurality of client terminals(Uehara par [0012] teaches a machine learning system includes a plurality of client terminals and an integration server) , the federated learning system performing federated learning, in which each of the plurality of client terminals learns a model using each of the learning data sets and the server repeats a process of integrating the model of each of the plurality of client terminals using the learning result(Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy), wherein the learning data set includes one or more data samples including a client ID specifying the client terminal, a first explanatory variable, and a first objective variable, (Uehara Fig.3 and par [0064] teaches client ID, imaging condition, learning result )
the federated learning system executes a first calculation process by the server and a second calculation process by each of the plurality of client terminals, and executes a first federated learning process that repeats a first learning process by each of the plurality of client terminals and a first integration process by the server until a first end condition is satisfied,( Uehara par [0072] teaches local LD model . Uehara par [0076] teaches the integration server receives each learning result and the imaging condition information from the plurality of clients. Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy)
in the first calculation process, the server obtains an input data set including one or more pieces of input data that includes the client ID specifying the client terminal, the first explanatory variable, and the first objective variable, and [calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets], (Uehara par [0084] teaches the master model learning management program of the integration server evaluates the inference accuracy of the master model candidate MMC using verification data for each of various imaging condition categories prepared in advance)
in the second calculation process, each of the plurality of client terminals [calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model for calculating the similarity between the data sample and the plurality of learning data sets], (Uehara par [0081] teaches the data of the local model LM as a learning results provided from each client to the integration server may be the weight parameter of the trained local model LM. Uehara par [0087] teaches the client CL1 performs the training of a local model LM11 using the local data LD11 as the learning data)
in the first learning process, each of the plurality of client terminals learns an individual analysis model for calculating a predicted value of the first objective variable from the first explanatory variable based on the individual analysis model, the first explanatory variable, the first objective variable (Uehara par [0081] teaches the data of the local model LM as a learning results provided from each client to the integration server may be the weight parameter of the trained local model LM. Uehara par [0092] teaches the learning results from clients are transmitted to the serve) , and a specific second similarity with a specific learning data set calculated by each of the plurality of client terminals by the calculation process (Uehara par [0075] teaches the client transmits a learning result of each local model LM and imaging condition information of the data used for the learning to the integration server) , and
in the first integration process, the server generates an analysis model for calculating the predicted value of the first objective variable from the first explanatory variable for the input data set by integrating a plurality of first learning results by the first learning process from the plurality of client terminals, using the first similarity. (Uehara par [0094] teaches the master model candidate creation unit integrates the learning results of the client clusters for each imaging condition category classified by the imaging condition classification unit to create a master model candidate MMck.)
Uehara fails to teach calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets; calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model for calculating the similarity between the data sample and the plurality of learning data sets;
However, Dimitriadis teaches calculates a first similarity between the input data set and the plurality of learning data sets by inputting the input data set into a similarity calculation model for calculating a similarity between the data sample and the plurality of learning data sets; calculates a second similarity between the data sample and the plurality of learning data sets by inputting the data sample into the similarity calculation model for calculating the similarity between the data sample and the plurality of learning data sets; (Dimitriadis par [0043] teaches at least one gradient quality metric is determined for each gradient of the plurality of gradients. In some examples, the determined gradient quality metrics include mean values of the gradients, variance values of the gradients, and/or training loss values of the gradients)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teaching of Uehara and Dimitriadis to achieve the claimed invention . One would have been motivated to make such combination to increase accuracy or performance of the model .( Dimitriadis par [0029])
As to claim 5, Uehara and Dimitriadis teach the federated learning system according to claim 4, wherein each of the plurality of client terminals executes the first learning process using the analysis model generated by the first integration process as the individual analysis model. (Uehara par [0100] teaches the integration server synchronizes the latest version of the master model with the local model LM on each client before each clients starts the training)
As to claim 6, Uehara and Dimitriadis teach the federated learning system according to claim 4, wherein the server executes inference by inputting the input data set into the analysis model when the first end condition is satisfied.( Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy)
As to claim 7, Uehara and Dimitriadis teach the federated learning system according to claim 1, wherein in the second calculation process, each of the plurality of client terminals calculates the second similarity by inputting a combination of the first explanatory variable and the first objective variable of the data sample into the similarity calculation model as a second explanatory variable. (Uehara par [0087] teaches the client performs the training of local model using local data as the learning data)
As to claim 8, Uehara and Dimitriadis teach the federated learning system according to claim 4, wherein in the second calculation process, each of the plurality of client terminals calculates a learning weight according to the specific second similarity, and in the first learning process, each of the plurality of client terminals learns the individual analysis model based on the individual analysis model, the first explanatory variable, the first objective variable, and a learning weight corresponding to the specific second similarity calculated by each of the plurality of client terminals by the second calculation process. (Uehara par [0075] teaches the client transmits a learning result of each local model LM and imaging condition information of the data used for the learning to the integration server)
As to claim 9, Uehara and Dimitriadis teach the federated learning system according to claim 1, wherein the federated learning system executes, before the first federated learning process, a second federated learning process in which a second learning process by each of the plurality of client terminals and a second integration process by the server are repeated until a second end condition is satisfied, (Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master model candidate MMC achieves a desired inference accuracy)
in the second learning process, each of the plurality of client terminals learns a similarity calculation model to be learned using a combination of the first explanatory variable and the first objective variable as a second explanatory variable and the client ID as a second objective variable . ( Uehara par [0087] teaches the client CL1 performs the training of a local model LM11 using the local data LD11 as the learning data), and
in the second integration process, the server integrates second learning results of the similarity calculation models to be learned from the plurality of client terminals by the second learning process and generates an integrated similarity calculation model obtained by integrating the similarity calculation models to be learned of the plurality of client terminals as the similarity calculation model. . (Uehara par [0094] teaches the master model candidate creation unit integrates the learning results of the client clusters for each imaging condition category classified by the imaging condition classification unit to create a master model candidate MMck.)
As to claim 10, Uehara and Dimitriadis teach the federated learning system according to claim 9, wherein in the second learning process, each of the plurality of client terminals sets the updated similarity calculation models to be learned as the similarity calculation model when the second end condition is satisfied. (Uehara par [0085] teaches the integration server repeats the learning in the client and the training of the master model candidate MMC for each identical imaging condition until each master
Claim 11 merely recites a method performed by the system of claim 1. Accordingly, Uehara and Dimitriadis teach every limitation of claim 11 as indicates in the above rejection of claim 1.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Martins. US Patent Application Publication 2023/0044035 , par [0004] discloses federated learning works in rounds. Tsuyuki., US Patent Application Publication 2022/0284061 A1, abstract, discloses a prediction unit executes a prediction by each of a plurality of trained models using test data. A similarity calculation unit calculates similarities between prediction results of the trained models as similarities of the trained models.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HIEN DUONG whose telephone number is (571)270-7335. The examiner can normally be reached Monday-Friday 8:00AM-5:00PM.
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, Viker Lamardo can be reached at 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 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.
/HIEN L DUONG/Primary Examiner, Art Unit 2147
Prosecution Insights