DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application is being examined under the pre-AIA first to invent provisions.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 06/02/2023 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Arguments
101 Rejection Arguments
Applicant asserts:
Applicant argues, on page 9-12, that “The Claims Integrate any Alleged Abstract Idea into a Practical Application”
Examiner response:
Examiner respectfully disagrees. Examiner interprets the generating step as an abstract idea. In addition, the additional elements of sending a model update and sending the training metadata is interpreted as extra solution activity mere data gathering. As states by MPEP 2106.04(d) “The courts have also identified limitations that did not integrate a judicial exception into a practical application…Adding insignificant extra-solution activity to the judicial exception” Therefore the abstract idea is not integrated into a practical application.
Applicant asserts:
Applicant argues, on page 12-13, that “The Claims Recite an Inventive Concept that Amounts to Significantly More than an Abstract Idea”
Examiner response:
Examiner respectfully disagrees. Examiner interprets the generating step as an abstract idea. In addition, the additional elements of sending a model update and sending the training metadata is interpreted as extra solution activity mere data gathering. As states by MPEP 2106.05 “Limitations that the courts have found not to be enough to qualify as "significantly more" when recited in a claim with a judicial exception include… Adding insignificant extra-solution activity to the judicial exception” Therefore the additional elements to not amount to significantly more when recited with the judicial exception.
103 Rejection Arguments
Applicant asserts:
Applicant argues, on page 14-15, that “Vandikas in view of Watanabe does not teach or suggest "generating, by the client device, training metadata using a trained local machine learning model and local validation data,"”
Examiner response:
Examiner respectfully disagrees. Using BRI, the limitation is interpreted as using a trained local machine learning model to generate/obtain training metadata. As stated in Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training.” The control signaling data as training metadata is generated by training/using the model from during or after training. Therefore, the prior art does teach the cited limitation.
Applicant asserts:
Applicant argues, on page 14-15, that “Vandikas in view of Watanabe does not teach or suggest the training metadata comprises data about the trained local machine learning model used to determine when to discontinue federated learning operations for training the global machine learning model”
Examiner response:
Examiner respectfully disagrees. Using BRI, the limitation is interpreted as the training metadata comprises data about the ML model to stop federated learning operations for training the global model. Federated learning operations is interpreted broadly as any operation. As stated in Vandikas Paragraph 0062, an indication of further training rounds or not is discontinuing federated learning operations. Therefore, the prior art does teach the cited limitation.
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 1-7 and 23-29 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
In reference to claim 1:
Step 1 - Is the claim to a process, machine, manufacture or composition of matter?
Yes, the claim is directed to a method
Step 2A Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon?
“generating, by the client device, training metadata using a trained local machine learning model and local validation data, wherein: the trained local machine learning model incorporates the model update data and global model data defining a global machine learning model, and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue federated learning operations for training the global machine learning model;” which is an abstract idea because it is directed to a mental process, an observation, evaluation, judgement, or opinion. The limitation as drafted, and under a broadest reasonable interpretation, can be performed in the human mind, or by a human using a pen and paper (MPEP 2106.04(a)(2)(Ill)(c)). For example, a person could generate training metadata wherein the person could analyze validation data and generate a number of training rounds before discontinuing federated learning operations.
Step 2A Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application?
“A computer-implemented method, comprising: sending, by a client device. model update data to a server;” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
“and sending, by the client device, the training metadata to the server.” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
The claim does not include additional elements that are integrated into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception?
“A computer-implemented method, comprising: sending, by a client device. model update data to a server;” (well-understood, routine, conventional MPEP 2106.05(d))
“and sending, by the client device, the training metadata to the server.” (well-understood, routine, conventional MPEP 2106.05(d))
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
In reference to claim 2:
Claim 2 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 3:
Claim 3 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 4:
Claim 4 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 5:
Claim 5 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 6:
Step 1 - Is the claim to a process, machine, manufacture or composition of matter?
Yes, the claim is directed to a process
Step 2A Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon?
“and processing, by the client device, local data with an updated global machine learning model based on the updated global model data to perform a task.” which is an abstract idea because it is directed to a mental process, an observation, evaluation, judgement, or opinion. The limitation as drafted, and under a broadest reasonable interpretation, can be performed in the human mind, or by a human using a pen and paper (MPEP 2106.04(a)(2)(Ill)(c)). For example, a person could process local data to perform a task/classification.
Step 2A Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application?
“The method of Claim 2, further comprising: receiving, by the client device, updated global model data from the server;” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
The claim does not include additional elements that are integrated into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception?
“The method of Claim 2, further comprising: receiving, by the client device, updated global model data from the server;” (well-understood, routine, conventional MPEP 2106.05(d))
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
In reference to claim 7:
Claim 7 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 23:
Step 1 - Is the claim to a process, machine, manufacture or composition of matter?
Yes, the claim is directed to a manufacture
Step 2A Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon?
“generate training metadata using a trained local machine learning model and local validation data, wherein: the trained local machine learning model incorporates the model update data and global model data defining a global machine learning model, and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue federated learning operations for training the global machine learning model;” which is an abstract idea because it is directed to a mental process, an observation, evaluation, judgement, or opinion. The limitation as drafted, and under a broadest reasonable interpretation, can be performed in the human mind, or by a human using a pen and paper (MPEP 2106.04(a)(2)(Ill)(c)). For example, a person could generate training metadata wherein the person could analyze validation data and generate a number of training rounds before discontinuing federated learning operations.
Step 2A Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application?
“A system comprising: memory comprising computer-executable instructions; and at least one processor configured to execute the computer-executable instructions in order to cause the system to:“ is merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“send model update data to a server;” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
“and send the training metadata to the server.” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
The claim does not include additional elements that are integrated into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception?
“A system comprising: memory comprising computer-executable instructions; and at least one processor configured to execute the computer-executable instructions in order to cause the system to:“ is merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“send model update data to a server;” (well-understood, routine, conventional MPEP 2106.05(d))
“and send the training metadata to the server.” (well-understood, routine, conventional MPEP 2106.05(d))
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
In reference to claim 24:
Claim 24 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 25:
Claim 25 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 26:
Claim 26 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 27:
Claim 27 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
In reference to claim 28:
Step 1 - Is the claim to a process, machine, manufacture or composition of matter?
Yes, the claim is directed to a manufacture
Step 2A Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon?
“and process local data with an updated global machine learning model based on the updated global model data to perform a task.” which is an abstract idea because it is directed to a mental process, an observation, evaluation, judgement, or opinion. The limitation as drafted, and under a broadest reasonable interpretation, can be performed in the human mind, or by a human using a pen and paper (MPEP 2106.04(a)(2)(Ill)(c)). For example, a person could process local data to perform a task/classification.
Step 2A Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application?
“The system of Claim 24, wherein the processor is further configured to cause the system to: updated global model data from the server;” (insignificant extra-solution activity mere data gathering MPEP 2106.05(g))
The claim does not include additional elements that are integrated into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception?
“The system of Claim 24, wherein the processor is further configured to cause the system to: updated global model data from the server;” (well-understood, routine, conventional MPEP 2106.05(d))
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
In reference to claim 29:
Claim 29 is directed to a judicial exception from claim(s) depended on and does not recite additional elements that integrate the judicial exception into a practical application and amount to significantly more than the judicial exception.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-5 and 23-27 are rejected under 35 U.S.C. 103 as being unpatentable over Konstantinos Vandikas et al; US 20220052925 A1 filed on Dec 6, 2019 (hereinafter “Vandikas”) in view of Shoichiro Watanabe et al; US 20230050708 A1 filed on Aug 16, 2021 (hereinafter “Watanabe”).
Regarding claim 1, Vandikas teaches A computer-implemented method, comprising: sending, by a client device, model update data to a server; (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Examiner notes that a client device (edge node 10-N) sends/transmits model update data (information 18-N) to a server (server 12))
generating, by the client device, training metadata using a trained local machine learning model and local validation data (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Vandikas Paragraph 0064; “Note that the accuracy of the local model as trained by an edge node may be with respect to a local test dataset at the edge node” Examiner notes that training metadata (control signaling that indicates an accuracy) is generated by the client device (edge node 10-N) using a trained local machine learning model (local model 14-N as seen in previous mapping) and local validation data (local test dataset 22-N))
wherein: the trained local machine learning model incorporates the model update data and global model data [defining a global machine learning model], (Vandikas Paragraph 0058; “the multi-node training information 20-1, 20-2, . . . 20-N may take the form of averaged weights.” Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Vandikas Paragraph 0068; “In some embodiments, the method includes training a local model 14-1 of network communication performance over one or more rounds of training at the edge node 10-1, based on a local training dataset 16-1 at the edge node 10-1 and based on multi-node training information 20-1 received in each round of training” Examiner notes that the trained local machine model (local model is trained) incorporates the model update data (local model incorporates information 18-N by generating it as a part of training) and global model data (multi-node training information))
and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue federated learning operations [for training the global machine learning model]; (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling … that indicates an accuracy of the local model as trained by the edge node through that round of training, that indicates whether another round of training is needed or desired at the edge node, … based on at least the accuracy of the local model as trained by the edge node through that round of training, whether one or more conditions are met for stopping training of the local model at the edge node.”)
and sending, by the client device, the training metadata to the server. (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Examiner notes that the client device (edge node) sends/transmits the training metadata (control signaling that indicates an accuracy) to the server (server 12))
Vandikas does not teach [wherein: the trained local machine learning model incorporates the model update data and] global model data defining a global machine learning model,
[and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue] federated learning operations for training the global machine learning model;
However, Watanabe does teach [wherein: the trained local machine learning model incorporates the model update data and] global model data defining a global machine learning model, (Watanabe Paragraph 0016; “Once the training nodes complete their local training, the results can be combined at a central server to generate a global model that captures the learned patterns of its constituent locally-trained models.” Examiner notes that global model data (training results) define a global machine learning model (combined at a central server to generate a global model))
[and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue] federated learning operations for training the global machine learning model; (Watanabe Paragraph 0027; “Federated learning module 115 may perform various functions associated with federated learning tasks, such as generating and distributing initial training algorithms to distributed computing nodes 130A-130N for training, combining results of the training process to generate trained global models” Examiner notes that global machine learning model (global models) is trained (training algorithms))
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas and Watanabe. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. One of ordinary skill would have motivation to combine Vandikas and Watanabe to improve the accuracy of the model by reducing or eliminating bias “present invention embodiments improve the field of machine learning by reducing or eliminating bias during the training phase of federated learning and/or other collaborative learning models, thereby enabling global models to be developed that are more accurate than models produced by conventional techniques.” (Watanabe Paragraph 0020).
Regarding claim 2, Vandikas teaches The method of Claim 1, further comprising: receiving, by the client device, the global model data from the server; (Vandikas Paragraph 0057; “the server 12 generates multi-node training information 20-1, 20-2, . . . 20-N based on the information 18-1, 18-2, . . . 18-N received from the edge nodes and transmits that multi-node training information towards the edge nodes.” Examiner notes that the client device (edge nodes) receive the global model data (multi-node training information) from the server (server 12 transmit data))
generating, by the client device, the local machine learning model based on the global model data; (Vandikas Paragraph 0059; “ Any given edge node may accordingly train its local model based on this multi-node training information.” Examiner notes that the client device (edge node) generates the local machine learning model (trains the local model) based on the global model data (multi-node training information))
and training, by the client device, the local machine learning model using local training data to generate the model update data. (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Examiner notes that the client device (edge node) trains the local machine learning model (each round of training of the local model) using local training data (local training dataset) to generate the model update data (information 18-N))
Regarding claim 3, Vandikas teaches The method of Claim 2, wherein the model update data and the training metadata are sent to the server synchronously. (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12” Vandikas Paragraph 0062; “after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Examiner notes that the model update data (information 18-N) and the training metadata (control signaling that indicates an accuracy) are sent to the server synchronously (transmits both information and control signaling as a part of each training round))
Regarding claim 4, Vandikas teaches The method of Claim 2, further comprising receiving, by the client device, updated global model data from the server prior to sending the training metadata to the server. (Vandikas Figure 2 and Paragraph 0076; “the multi-node training information 20-1 includes a combination of local updates that the other edge nodes respectively made to local models at the other edge nodes.” Examiner notes that the client device (edge node) receives updated global model data (multi-node training information in step 200 received after first round of training) from the server prior to sending the training metadata (control signaling in step 210) to the server)
PNG
media_image1.png
569
745
media_image1.png
Greyscale
Regarding claim 5, Vandikas teaches The method of Claim 2, wherein: the training metadata comprises one or more of:(i) a first accuracy value based on testing the trained local machine learning model with the local training data, or(ii) a loss value associated with training the local machine learning model based on the local training data; And the training metadata comprises a second accuracy value associated with testing the trained local machine learning model using the local validation data. (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training,” Vandikas Paragraph 0064; “Note that the accuracy of the local model as trained by an edge node may be with respect to a local test dataset at the edge node” Examiner notes that training metadata comprises a first accuracy (accuracy of the local model) based on testing the trained local machine learning model (after or part of training the model accuracy is transmitted) with the local training data (local test dataset); Examiner interprets a second accuracy is included in the one or more of the training metadata)
Regarding claim 23, Vandikas teaches A system comprising: memory comprising computer-executable instructions; and at least one processor configured to execute the computer-executable instructions in order to cause the system to: (Vandikas Paragraph 0133; “The edge node comprises one or more processors and a memory, the memory containing instructions executable by the one or more processors whereby the edge node is configured to perform the method shown in any of FIGS. 2-4.” Vandikas Paragraph 0137; “the server comprising one or more processors and a memory, the memory containing instructions executable by the one or more processors whereby the server is configured to perform the method shown in any of FIGS. 5-7.”)
send model update data to a server; (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Examiner notes that a client device (edge node 10-N) sends/transmits model update data (information 18-N) to a server (server 12))
generate training metadata using a trained local machine learning model and local validation data (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Vandikas Paragraph 0064; “Note that the accuracy of the local model as trained by an edge node may be with respect to a local test dataset at the edge node” Examiner notes that training metadata (control signaling that indicates an accuracy) is generated by the client device (edge node 10-N) using a trained local machine learning model (local model 14-N as seen in previous mapping) and local validation data (local test dataset 22-N))
wherein: the trained local machine learning model incorporates the model update data and global model data [defining a global machine learning model], (Vandikas Paragraph 0058; “the multi-node training information 20-1, 20-2, . . . 20-N may take the form of averaged weights.” Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Vandikas Paragraph 0068; “In some embodiments, the method includes training a local model 14-1 of network communication performance over one or more rounds of training at the edge node 10-1, based on a local training dataset 16-1 at the edge node 10-1 and based on multi-node training information 20-1 received in each round of training” Examiner notes that the trained local machine model (local model is trained) incorporates the model update data (local model incorporates information 18-N by generating it as a part of training) and global model data (multi-node training information))
and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue federated learning operations [for training the global machine learning model]; (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling … that indicates an accuracy of the local model as trained by the edge node through that round of training, that indicates whether another round of training is needed or desired at the edge node, … based on at least the accuracy of the local model as trained by the edge node through that round of training, whether one or more conditions are met for stopping training of the local model at the edge node.”)
and send the training metadata to the server. (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Examiner notes that the client device (edge node) sends/transmits the training metadata (control signaling that indicates an accuracy) to the server (server 12))
Vandikas does not teach [wherein: the trained local machine learning model incorporates the model update data and] global model data defining a global machine learning model,
[and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue] federated learning operations for training the global machine learning model;
However, Watanabe does teach [wherein: the trained local machine learning model incorporates the model update data and] global model data defining a global machine learning model, (Watanabe Paragraph 0016; “Once the training nodes complete their local training, the results can be combined at a central server to generate a global model that captures the learned patterns of its constituent locally-trained models.” Examiner notes that global model data (training results) define a global machine learning model (combined at a central server to generate a global model))
[and the training metadata comprises data about the trained local machine learning model used to determine when to discontinue] federated learning operations for training the global machine learning model; (Watanabe Paragraph 0027; “Federated learning module 115 may perform various functions associated with federated learning tasks, such as generating and distributing initial training algorithms to distributed computing nodes 130A-130N for training, combining results of the training process to generate trained global models” Examiner notes that global machine learning model (global models) is trained (training algorithms))
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas and Watanabe. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. One of ordinary skill would have motivation to combine Vandikas and Watanabe to improve the accuracy of the model by reducing or eliminating bias “present invention embodiments improve the field of machine learning by reducing or eliminating bias during the training phase of federated learning and/or other collaborative learning models, thereby enabling global models to be developed that are more accurate than models produced by conventional techniques.” (Watanabe Paragraph 0020).
Regarding claim 24, Vandikas teaches The system of Claim 23, wherein the processor is further configured to cause the system to: receive the global model data from the server; (Vandikas Paragraph 0057; “the server 12 generates multi-node training information 20-1, 20-2, . . . 20-N based on the information 18-1, 18-2, . . . 18-N received from the edge nodes and transmits that multi-node training information towards the edge nodes.” Examiner notes that the client device (edge nodes) receive the global model data (multi-node training information) from the server (server 12 transmit data))
Generate the local machine learning model based on the global model data; (Vandikas Paragraph 0059; “ Any given edge node may accordingly train its local model based on this multi-node training information.” Examiner notes that the client device (edge node) generates the local machine learning model (trains the local model) based on the global model data (multi-node training information))
and train the local machine learning model using local training data to generate the model update data. (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12 about its local model 14-1, 14-2, . . . 14-N as trained (based on its local training dataset) through that round of training.” Examiner notes that the client device (edge node) trains the local machine learning model (each round of training of the local model) using local training data (local training dataset) to generate the model update data (information 18-N))
Regarding claim 25, Vandikas teaches The system of Claim 24, wherein the model update data and the training metadata are sent to the server synchronously. (Vandikas Paragraph 0057; “each edge node 10-1, 10-2, . . . 10-N, after or as part of each round of training, transmits information 18-1, 18-2, . . . 18-N to the server 12” Vandikas Paragraph 0062; “after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training” Examiner notes that the model update data (information 18-N) and the training metadata (control signaling that indicates an accuracy) are sent to the server synchronously (transmits both information and control signaling as a part of each training round))
Regarding claim 26, Vandikas teaches The system of Claim 24, wherein the processor is configured to cause the system to receive updated global model data from the server prior to sending the training metadata to the server. (Vandikas Figure 2 and Paragraph 0076; “the multi-node training information 20-1 includes a combination of local updates that the other edge nodes respectively made to local models at the other edge nodes.” Examiner notes that the client device (edge node) receives updated global model data (multi-node training information in step 200 received after first round of training) from the server prior to sending the training metadata (control signaling in step 210) to the server)
PNG
media_image1.png
569
745
media_image1.png
Greyscale
Regarding claim 27, Vandikas teaches The system of Claim 24, wherein: the training metadata comprises one or more of:(i) a first accuracy value based on testing the trained local machine learning model with the local training data, or(ii) a loss value associated with training the local machine learning model based on the local training data; And the training metadata comprises a second accuracy value associated with testing the trained local machine learning model using the local validation data. (Vandikas Paragraph 0062; “an edge node, after or as part of each round of training, transmits control signaling (e.g., to the server 12 or other edge node(s)) that indicates an accuracy of the local model as trained by the edge node through that round of training,” Vandikas Paragraph 0064; “Note that the accuracy of the local model as trained by an edge node may be with respect to a local test dataset at the edge node” Examiner notes that training metadata comprises a first accuracy (accuracy of the local model) based on testing the trained local machine learning model (after or part of training the model accuracy is transmitted) with the local training data (local test dataset); Examiner interprets a second accuracy is included in the one or more of the training metadata)
Claim(s) 6-7 and 28-29 are rejected under 35 U.S.C. 103 as being unpatentable over Konstantinos Vandikas et al; US 20220052925 A1 filed on Dec 6, 2019 (hereinafter “Vandikas”) in view of Shoichiro Watanabe et al; US 20230050708 A1 filed on Aug 16, 2021 (hereinafter “Watanabe”) in further view of Ravindra Patil et al; US 20230252305 A1 filed on Jul 8, 2021 (hereinafter “Patil”).
Regarding claim 6, Vandikas teaches The method of Claim 2, further comprising: receiving, by the client device, updated global model data from the server; (Vandikas Figure 2 and Paragraph 0076; “the multi-node training information 20-1 includes a combination of local updates that the other edge nodes respectively made to local models at the other edge nodes.” Examiner notes that the client device (edge node) receives updated global model data (multi-node training information in step 200 received after first round of training) from the server)
Vandikas does not teach And processing, by the client device, local data with an updated global machine learning model based on the updated global model data to perform a task.
However, Patil does teach And processing, by the client device, local data with an updated global machine learning model based on the updated global model data to perform a task. (Patil Paragraph 0029; “there is a computer implemented method 300 for use in training a model to perform a task on (e.g. process) medical data using a distributed machine learning process whereby a global model is updated based on training performed on local copies of the model at a plurality of clinical sites.” Patil Paragraph 0041; “The model may take as input one or more of the types of medical data described above and perform a task on the medical data. The task may comprise, for example, a classification task or a segmentation task.” Examiner notes that local data (input of medical data) is processed by the client device with an updated global machine learning model (global model updated through training) based on the updated global model data (local copies of the model of a plurality of clinical sites) to perform a task)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas, Watanabe, and Patil. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. Patil teaches a method of training a model to perform a task on medical data using distributed machine learning process. One of ordinary skill would have motivation to combine Vandikas, Watanabe, and Patil to improve models for specific tasks through using appropriate data “metadata related to the quality of the training performed at each site may be used when combining the local updates into an update for the global model. In this way, different local updates may be given different significances (e.g. through the use of weightings) dependent on the quality of the training performed at the respective clinical site. This can result in improved training, resulting in improved models and thus improved clinical outcomes for clinical processes that use the models. Since the model is trained on data from different sites, there may be irregularities in the data, and this can lead to bias and model drift. By considering appropriate metadata while merging the weights, model drift may be avoided, leading to a better quality model.” (Patil Paragraph 0006).
Regarding claim 7, Vandikas does not teach The method of Claim 6, wherein the task comprises one of: image classification based on local image data; sound classification based on local sound data; or authentication based on local biometric data.
However, Patil does teach The method of Claim 6, wherein the task comprises one of: image classification based on local image data; sound classification based on local sound data; or authentication based on local biometric data. (Patil Paragraph 0041; “the model may output, for example, a segmentation of the medical image, a location of a feature of interest in the medical image, or a diagnosis based on the medical image.” Examiner notes that the task comprises image classification (diagnosis) based on local image data (medical image))
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas, Watanabe, and Patil. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. Patil teaches a method of training a model to perform a task on medical data using distributed machine learning process. One of ordinary skill would have motivation to combine Vandikas, Watanabe, and Patil to improve models for specific tasks through using appropriate data “metadata related to the quality of the training performed at each site may be used when combining the local updates into an update for the global model. In this way, different local updates may be given different significances (e.g. through the use of weightings) dependent on the quality of the training performed at the respective clinical site. This can result in improved training, resulting in improved models and thus improved clinical outcomes for clinical processes that use the models. Since the model is trained on data from different sites, there may be irregularities in the data, and this can lead to bias and model drift. By considering appropriate metadata while merging the weights, model drift may be avoided, leading to a better quality model.” (Patil Paragraph 0006).
Regarding claim 28, Vandikas teaches The system of Claim 24, wherein the processor is further configured to cause the system to: receive updated global model data from the server; (Vandikas Figure 2 and Paragraph 0076; “the multi-node training information 20-1 includes a combination of local updates that the other edge nodes respectively made to local models at the other edge nodes.” Examiner notes that the client device (edge node) receives updated global model data (multi-node training information in step 200 received after first round of training) from the server)
Vandikas does not teach And process local data with an updated global machine learning model based on the updated global model data to perform a task.
However, Patil does teach And process local data with an updated global machine learning model based on the updated global model data to perform a task. (Patil Paragraph 0029; “there is a computer implemented method 300 for use in training a model to perform a task on (e.g. process) medical data using a distributed machine learning process whereby a global model is updated based on training performed on local copies of the model at a plurality of clinical sites.” Patil Paragraph 0041; “The model may take as input one or more of the types of medical data described above and perform a task on the medical data. The task may comprise, for example, a classification task or a segmentation task.” Examiner notes that local data (input of medical data) is processed by the client device with an updated global machine learning model (global model updated through training) based on the updated global model data (local copies of the model of a plurality of clinical sites) to perform a task)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas, Watanabe, and Patil. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. Patil teaches a method of training a model to perform a task on medical data using distributed machine learning process. One of ordinary skill would have motivation to combine Vandikas, Watanabe, and Patil to improve models for specific tasks through using appropriate data “metadata related to the quality of the training performed at each site may be used when combining the local updates into an update for the global model. In this way, different local updates may be given different significances (e.g. through the use of weightings) dependent on the quality of the training performed at the respective clinical site. This can result in improved training, resulting in improved models and thus improved clinical outcomes for clinical processes that use the models. Since the model is trained on data from different sites, there may be irregularities in the data, and this can lead to bias and model drift. By considering appropriate metadata while merging the weights, model drift may be avoided, leading to a better quality model.” (Patil Paragraph 0006).
Regarding claim 29, Vandikas does not teach The system of Claim 28, wherein the task comprises one of: image classification based on local image data; sound classification based on local sound data; or authentication based on local biometric data.
However, Patil does teach The system of Claim 28, wherein the task comprises one of: image classification based on local image data; sound classification based on local sound data; or authentication based on local biometric data. (Patil Paragraph 0041; “the model may output, for example, a segmentation of the medical image, a location of a feature of interest in the medical image, or a diagnosis based on the medical image.” Examiner notes that the task comprises image classification (diagnosis) based on local image data (medical image))
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Vandikas, Watanabe, and Patil. Vandikas teaches a method of using federated learning to predict network communication performance. Watanabe teaches a system of training a federated learning model. Patil teaches a method of training a model to perform a task on medical data using distributed machine learning process. One of ordinary skill would have motivation to combine Vandikas, Watanabe, and Patil to improve models for specific tasks through using appropriate data “metadata related to the quality of the training performed at each site may be used when combining the local updates into an update for the global model. In this way, different local updates may be given different significances (e.g. through the use of weightings) dependent on the quality of the training performed at the respective clinical site. This can result in improved training, resulting in improved models and thus improved clinical outcomes for clinical processes that use the models. Since the model is trained on data from different sites, there may be irregularities in the data, and this can lead to bias and model drift. By considering appropriate metadata while merging the weights, model drift may be avoided, leading to a better quality model.” (Patil Paragraph 0006).
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 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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DANIEL DUC TRAN whose telephone number is (571)272-6870. The examiner can normally be reached Mon-Fri 8:00-5:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, 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.
/D.D.T./Examiner, Art Unit 2147
/VIKER A LAMARDO/Supervisory Patent Examiner, Art Unit 2147
Prosecution Insights