CTNF 18/512,195 CTNF 95804 DETAILED ACTION This action is in response to the claims filed 11/17/2023 for Application number 18/512,195 which claims foreign priority to GB2214033.9 filed 09/26/2022. Claims 1-20 are currently pending. Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Priority 02-26 AIA Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55. Information Disclosure Statement The information disclosure statements (IDS) submitted on 11/17/2023 and 03/28/2024 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner. Claim Objections 07-29-01 AIA Claim 13 is objected to because of the following informalities: The claim recites "client devices (312)" which should read "client devices" . Appropriate correction is required. Claim Rejections - 35 USC § 101 07-04-01 AIA 07-04 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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Regarding claim 1 , Step 1 Analysis: Claim 1 is directed to a process, which falls within one of the four statutory categories. Step 2A Prong 1 Analysis: Claim 1 recites, in part, The limitations of: defining, at a server, a Bayesian hierarchical model which links a global random variable with a plurality of local random variables, one for each of the plurality of client devices can be considered to be an evaluation in the human mind, to determine an updated global parameter can be considered to be an evaluation in the human mind These limitations as drafted, are processes that, under broadest reasonable interpretation, covers performance of the limitation in the mind or with the aid of pen and paper which falls within the “Mental Processes” grouping of abstract ideas. The limitations of: wherein the Bayesian hierarchical model comprises a posterior distribution which is suitable for predicting the likelihood of the global random variable and the local random variables given each individual dataset at the plurality of client devices can be considered to be a mathematical calculation approximating, at the server, the posterior distribution [using a global ML model and a plurality of local ML models], wherein the global ML model is parameterised by a global parameter which is updated at the server, each of the plurality of local ML models is associated with one of the plurality of client devices and each local ML model is parameterised by a local parameter which is updated at the client device which is associated with the local ML model can be considered to be a mathematical calculation These limitations as drafted, are processes that, under broadest reasonable interpretation, covers mathematical calculations which falls within the “Mathematical Concepts” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. Step 2A Prong 2 Analysis: This judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements – “for use by a plurality of client devices”, “using a global ML model and a plurality of local ML models”, and “training, at the server, the global ML model using each of the received updated local parameters [to determine an updated global parameter], wherein during training of the global model, each local parameter is fixed.”. Thus, these elements in the claim are recited at a high level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP 2106.05(f). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim further recites: sending, from the server, the global parameter to a predetermined number of the plurality of client devices; receiving, at the server from each of the number of the plurality of client devices, an updated local parameter, wherein each updated local parameter has been determined by training, on the client device, the local ML model using a local dataset, and wherein during training of the local ML model, the global parameter is fixed. These limitations are insignificant extra-solution activities. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim as a whole is directed to an abstract idea. Step 2B Analysis: The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of utilizing client devices, a global ML model, and a local ML model to perform the steps of the claimed process amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Furthermore, the limitations of sending, from the server, the global parameter to a predetermined number of the plurality of client devices; receiving, at the server from each of the number of the plurality of client devices, an updated local parameter, wherein each updated local parameter has been determined by training, on the client device, the local ML model using a local dataset, and wherein during training of the local ML model, the global parameter is fixed are well-understood, routine, and conventional, as evidenced by MPEP §2106.05(d)(II)(I), “receiving or transmitting data over a network”. These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and does not amount to significantly more. Even when considered in combination, these additional elements amount to mere instructions to apply the exception using generic computer components and insignificant extra-solution activity, which cannot provide an inventive concept. The claim is not patent eligible. Regarding claim 2 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein at least some of the local ML models and/or global ML model have different structures. This limitation amounts to more specifics of the judicial exception identified in the rejection of claim 1 above. The claim does not include any additional elements that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible . Regarding claim 3 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein training the global ML model comprises optimising using a regularization term which penalises deviation between the updated global parameter and the global parameter which was sent to the client devices. This claim recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception. The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible. Regarding claim 4 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein training the global ML model comprises optimising using a regularization term which penalises deviation between the updated global parameter and each of the local parameters received from the plurality of client devices. This claim recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception. The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible. Regarding claim 5 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein approximating the posterior distribution comprises using a Normal-Inverse-Wishart model as the global ML model and using a global mean parameter and a global covariance parameter as the global parameter and using a mixture of two Gaussian functions as the local ML model and a local mean parameter as the local parameter. This limitation amounts to more specifics of the judicial exception identified in the rejection of claim 1 above. The claim does not include any additional elements that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible. Regarding claim 6 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein approximating the posterior distribution comprises using a mixture model which comprises multiple different prototypes and each prototype is associated with a separate global random variable. This limitation amounts to more specifics of the judicial exception identified in the rejection of claim 1 above. The claim does not include any additional elements that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible. Regarding claim 7 , the rejection of claim 6 is further incorporated, and further, the claim recites: further comprising using a product of multiple multivariate normal distributions as the global model and using variational parameters as the global parameter and using one of the multiple multivariate normal distributions as the local ML model and a local mean parameter as the local parameter. This claim recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception. The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible. Regarding claim 8 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein training, on the client device, comprises optimising using a loss function to fit each local parameter to the local dataset. This claim recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception. The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible. Regarding claim 9 , the rejection of claim 1 is further incorporated, and further, the claim recites: wherein training, on the client device, comprises optimising using a regularisation term which penalises deviation between each updated local parameter and a previous local parameter. This claim recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception. The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible. Regarding claim 10 , Step 1 Analysis: Claim 10 is directed to a process, which falls within one of the four statutory categories. Step 2A Prong 1 Analysis: Claim 10 recites, in part, The limitation of: optimising, at the client device, a local parameter using the received global parameter, by applying a regularisation term which penalises deviation between the optimised local parameter and the received global parameter, This limitation as drafted, is a process that, under broadest reasonable interpretation, covers mathematical calculations which falls within the “Mathematical Concepts” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. Step 2A Prong 2 Analysis: This judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements – “using a client device” and “using a global machine learning, ML, model”. Thus, these elements in the claim are recited at a high level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP 2106.05(f). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim further recites: receiving, at the client device from the server, a global parameter for the trained global ML model outputting the optimized local parameter as the personalized model. These limitations are insignificant extra-solution activities. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim as a whole is directed to an abstract idea. Step 2B Analysis: The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of utilizing a client device and global ML model to perform the steps of the claimed process amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Furthermore, the limitations of receiving, at the client device from the server, a global parameter for the trained global ML model and outputting the optimized local parameter as the personalized model. are well-understood, routine, and conventional, as evidenced by MPEP §2106.05(d)(II)(I), “receiving or transmitting data over a network”. These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and does not amount to significantly more. Even when considered in combination, these additional elements amount to mere instructions to apply the exception using generic computer components and insignificant extra-solution activity, which cannot provide an inventive concept. The claim is not patent eligible. Regarding claim 11 , the rejection of claim 10 is further incorporated, and further, the claim recites: obtaining a set of personal data; This limitation is an insignificant extra-solution activity and thus the judicial exception is not integrated into a practical application. Additionally, the limitation of wherein optimising the local parameter using the received global parameter comprises optimising the local parameter using the received global parameter, by applying a loss function over the set of personal data recites additional mathematical calculations in addition to the judicial exception identified in the rejection of claim 10 above. The claim as a whole is directed to an abstract idea. The claim does not include any additional elements that amount to significantly more than the judicial exception. The limitation of obtaining a set of personal data is just a nominal or tangential addition to the claim, and is also well-understood, routine and conventional as evidenced by MPEP §2106.05(d)(II)(I), “receiving or transmitting data over a network”. This limitation therefore remains insignificant extra-solution activity even upon reconsideration, and does not amount to significantly more. Even when considered in combination, this additional element represents an insignificant extra-solution activity which cannot provide an inventive concept. The claim is not patent eligible. Regarding claim 12 , the rejection of claim 10 is further incorporated, and further, the claim recites: receiving an input; This limitation is an insignificant extra-solution activity and thus the judicial exception is not integrated into a practical application. Additionally, the limitation of predicting, using the personalised model, an output based on the received input recites additional mental steps in addition to the judicial exception identified in the rejection of claim 10 above. The claim as a whole is directed to an abstract idea. The claim does not include any additional elements that amount to significantly more than the judicial exception. The limitation of receiving an input is just a nominal or tangential addition to the claim, and is also well-understood, routine and conventional as evidenced by MPEP §2106.05(d)(II)(I), “receiving or transmitting data over a network”. This limitation therefore remains insignificant extra-solution activity even upon reconsideration, and does not amount to significantly more. Even when considered in combination, this additional element represents an insignificant extra-solution activity which cannot provide an inventive concept. The claim is not patent eligible. Regarding Claim 13 , it recites features similar to claim 1 and is rejected for at least the same reasons therein. Regarding Claims 14-19 , they recite features similar to claims 2-6 and 9 and is rejected for at least the same reasons therein. Claim 20 recites features similar to claim 1 and is rejected for at least the same reasons therein. Claim 20 additionally requires analysis for “A non-transitory storage media storing a computer program that, when executed by at least one processor, causes the at least one processor to perform the method of claim 1.” However, this is an additional element which amounts to mere instructions to apply the judicial exception using a generic computer component. Please see MPEP 2106.05(f). Claim Rejections - 35 USC § 102 07-06 AIA 15-10-15 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. 07-07-aia AIA 07-07 The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action: A person shall be entitled to a patent unless – 07-12-aia AIA (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. 07-15-03-aia AIA Claim s 10-12 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Zhang et al. ("Personalized Federated Learning via Variational Bayesian Inference" cited by Applicant in the IDS filed 03/28/2024, hereinafter "Zhang") . Regarding claim 10 , Zhang teaches A method for generating, using a client device, a personalised model using a global machine learning, ML, model which has been trained at a server, the method comprising: receiving, at the client device from the server, a global parameter for the trained global ML model (“In Fig. 1(b), we show the evolution of the distribution training, where the global model (GM) serves as the prior of the personalized models (PMs). The figure presents that the trained distribution for each PM lies between the GM and PM without communication, which balances the GM and extreme PM without communication.” [pg. 3, 2., ¶1]) ; optimising, at the client device, a local parameter using the received global parameter, by applying a regularisation term which penalises deviation between the optimised local parameter and the received global parameter (“By minimizing the sum of KL divergence, the global model can find the closest distribution in Qw to the clients’ distribution. Note that in the regularization term for clients, we replace prior distribution with global distribution” [pg. 4, left col, ¶2]) , and outputting the optimised local parameter as the personalised model (“In this section, we present the personalized Bayesian fed erated learning model with Gaussian distribution. In our model, each client owns its BNN, where each parameter (weight or bias) denotes a Gaussian distribution… As shown in Fig. 1(a), at each iteration, the clients upload their distributions (i.e., mean and variance) to the server and then download the aggregated distribution from the server.” [pg. 3, 2., ¶1]) . Regarding claim 11 , Zhang teaches The method of claim 10, further comprising: obtaining a set of personal data (“Each client uses the aggregated global distribution as prior distribution and updates its personal distribution by balancing the construction error over its personal data and the KL di vergence with aggregated global distribution.” [pg. 9, §6. Conclusions, 1]) , wherein optimising the local parameter using the received global parameter comprises optimising the local parameter using the received global parameter, by applying a loss function over the set of personal data (“According to Theorem 1,we notice that with the increase of ζ, the approximation error in the upper bound decreases, so does the upper bound. But from the formulation of the proposed optimization problem, we know the increase of ζ will decrease the degree of personalization. So a suitable choice of ζ is necessary to balance the degree of personalization and the value of the global upper bound.” [pg. 6, bottom left col]) . Regarding claim 12 , Zhang teaches The method of claim 10, further comprising: receiving an input; and predicting, using the personalised model, an output based on the received input. (“On small, medium and large datasets of MNIST (“receiving an input”) , the personalized model of our algorithm outperforms other SOTA comparison algorithms by 1.25%, 1.78% and 0.52%, while the global model out performs other SOTA models by 2.79%, 1.67% and 1.97%, respectively.” [pg. 7, §5.3., ¶1]) Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA 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. 07-23-aia AIA The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. 07-21-aia AIA Claim s 1-4, 6-9, 13-15, and 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang in view of Chen et al. ("Self-Aware Personalized Federated Learning" cited by Applicant in the IDS filed 03/28/2024, hereinafter "Chen") . Regarding claim 1 , Zhang teaches A method for training, using federated learning, a global machine learning, ML, model for use by a plurality of client devices (Abstract) , the method comprising: approximating, at the server, the posterior distribution using a global ML model and a plurality of local ML models (“The optimization problem aims to find the closest distribution to the posterior distribution in the variational family of distributions Q” [pg. 3, right col, bottom para]) , wherein the global ML model is parameterised by a global parameter which is updated at the server (“where w(θ)and qi(θ)denote the global distribution and the local distribution for the i-th client to be optimized, respectively, Qw and Qi denote the family of distributions for global parameter and the i-th client, respectively” [pg. 4, left col ¶2]) , each of the plurality of local ML models is associated with one of the plurality of client devices and each local ML model is parameterised by a local parameter which is updated at the client device which is associated with the local ML model (“The local model encourages clients to find a local distribution that balances the construction error over its private data and the KL divergence with global distribution.” [pg. 2, top left col]) ; sending, from the server, the global parameter to a predetermined number of the plurality of client devices (See Figure 1, pg. 2, “Left: System diagram. Each client uploads its updated distribution to the server and t hen downloads the aggregated global distribution from the server ”) ; receiving, at the server from each of the number of the plurality of client devices, an updated local parameter, wherein each updated local parameter has been determined by training, on the client device, the local ML model using a local dataset, and wherein during training of the local ML model, the global parameter is fixed (“In our model, each client owns its BNN, where each parameter (weight or bias) denotes a Gaussian distribution… As shown in Fig. 1(a), at each iteration, the clients upload their distributions (i.e., mean and variance) to the server and then download the aggregated distribution from the server.” [pg. 3, 2., ¶1; See further pg. 1, right col, “Unlike traditional BNNs, we do not assume a prior distribution for each parameter on the end devices; instead, the trained global distribution is used as the prior distribution. ”]); and training, at the server, the global ML model using each of the received updated local parameters to determine an updated global parameter, wherein during training of the global model, each local parameter is fixed (“After R iterations, the clients upload the localized global models to the server. On the server side, since clients are sometimes silent, we assume that a random set St ∈ {1,...,N} with size S is available for the server. After receiving the uploaded models, the server uses the average of clients in St to update the global model.” [pg. 5, right col bottom para – pg. 6, left col, top para]) . However fails to explicitly teach defining, at a server, a Bayesian hierarchical model which links a global random variable with a plurality of local random variables, one for each of the plurality of client devices, wherein the Bayesian hierarchical model comprises a posterior distribution which is suitable for predicting the likelihood of the global random variable and the local random variables given each individual dataset at the plurality of client devices; Chen teaches defining, at a server, a Bayesian hierarchical model which links a global random variable with a plurality of local random variables, one for each of the plurality of client devices, wherein the Bayesian hierarchical model comprises a posterior distribution which is suitable for predicting the likelihood of the global random variable and the local random variables given each individual dataset at the plurality of client devices; ( “From the server’s perspective, it is postulated that data z1,...,zM are generated from the following two-layer Bayesian hierarchical model:.. for all clients with indices m= 1,...,M. Here, σ2 0 is a constant, and θ0 is a hyperparameter with a non informative flat prior (denoted by π0). The above model represents both the connections and heterogeneity across clients. In particular, each client’s data are distributed according to a client-specific parameter (θm), which follows a distribution decided by a parent parameter (θ0) (links a global random variable with a plurality of local random variables) … Under the above model assumption, the parent parameter θ0 that represents the global model has a posterior distribution” [pg. 3, 2.1, ¶1]). It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Zhang’s teachings in order to implement the hierarchical Bayesian model of Chen. One would have been motivated to make this modification in order to achieve improved personalization performance when compared to existing techniques. [Abstract, Chen] Regarding claim 2 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein at least some of the local ML models and/or global ML model have different structures. (“Suppose each client has the same neural network architecture, i.e., a fully-connected Deep Neural Network (DNN), but has different parameters due to the fact that their data are non-i.i.d.” [pg. 3, §2., ¶3]) Regarding claim 3 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein training the global ML model comprises optimising using a regularization term which penalises deviation between the updated global parameter and the global parameter which was sent to the client devices. (“By minimizing the sum of KL divergence, the global model can find the closest distribution in Qw to the clients’ distribution. Note that in the regularization term for clients, we replace prior distribution with global distribution” [pg. 4, left col, ¶2]) . Regarding claim 4 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein training the global ML model comprises optimising using a regularization term which penalises deviation between the updated global parameter and each of the local parameters received from the plurality of client devices. (“By minimizing the sum of KL divergence, the global model can find the closest distribution in Qw to the clients’ distribution. Note that in the regularization term for clients, we replace prior distribution with global distribution” [pg. 4, left col, ¶2]) . Regarding claim 6 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein approximating the posterior distribution comprises using a mixture model which comprises multiple different prototypes and each prototype is associated with a separate global random variable. (“Recently, the Bayesian neural network (BNN) is introduced into FL to address the model overfitting by representing all network parameters in the global model with probability distributions” [pg. 1, right col, ¶2; See further, pg. 4, top right col, “Armed with the above distributions, we can obtain the KL divergences of the two Gaussian distributions qi(θ)and w(θ) as follows”]) Regarding claim 7 , Zhang/Chen teaches The method of claim 6, Zhang teaches further comprising using a product of multiple multivariate normal distributions as the global model (“Besides, it’s common to assume that the distribution satisfies mean field decomposition, that is, the joint distribution equals to the product of each parameter’s distribution.” [pg. 4, left col, bottom para]) and using variational parameters as the global parameter and using one of the multiple multivariate normal distributions as the local ML model and a local mean parameter as the local parameter. (“Let w ⋆ (θ) be the optimal variational solution of the problem and ˆqi(θ) be its corresponding variational solution for the i-th client’s subproblem” [pg. 5, left col]) Regarding claim 8 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein training, on the client device, comprises optimising using a loss function to fit each local parameter to the local dataset. (“According to Theorem 1,we notice that with the increase of ζ, the approximation error in the upper bound decreases, so does the upper bound. But from the formulation of the proposed optimization problem, we know the increase of ζ will decrease the degree of personalization. So a suitable choice of ζ is necessary to balance the degree of personalization and the value of the global upper bound.” [pg. 6, bottom left col]) Regarding claim 9 , Zhang/Chen teaches The method of claim 1, Zhang teaches wherein training, on the client device, comprises optimising using a regularisation term which penalises deviation between each updated local parameter and a previous local parameter. (“By minimizing the sum of KL divergence, the global model can find the closest distribution in Qw to the clients’ distribution. Note that in the regularization term for clients, we replace prior distribution with global distribution” [pg. 4, left col, ¶2]) . Regarding claim 13 , it is substantially similar to claim 1 respectively, and is rejected in the same manner, the same art, and reasoning applying. Regarding claims 14-15 and 17-19 , they are substantially similar to claims 3-4, 6-7 and 9 respectively, and are rejected in the same manner, the same art, and reasoning applying. Claim 20 recites features similar to claim 1 and is rejected for at least the same reasons therein. Claim 20 additionally requires A non-transitory storage media storing a computer program that, when executed by at least one processor, causes the at least one processor to perform the method of claim 1. (“We did all experiments in this paper using servers with two GPUs (NVIDIA Tesla P100 with 16GB memory), two CPUs (each with 22 cores, Intel(R) Xeon(R) Gold 6152 CPU @2.10GHz), and 192 GB memory” [pg. 7, left col, bottom para]) 07-21-aia AIA Claim s 5 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang in view of Chen and further in view of Benhamou et al. ("BCMA-ESII: revisiting Bayesian CMA-ES", hereinafter "Benhamou") . Regarding claim 5 , Zhang/Chen teaches The method of claim 1, however fails to explicitly teach wherein approximating the posterior distribution comprises using a Normal-Inverse-Wishart model as the global ML model and using a global mean parameter and a global covariance parameter as the global parameter and using a mixture of two Gaussian functions as the local ML model and a local mean parameter as the local parameter. Benhamou teaches wherein approximating the posterior distribution comprises using a Normal-Inverse-Wishart model as the global ML model and using a global mean parameter and a global covariance parameter as the global parameter and using a mixture of two Gaussian functions as the local ML model and a local mean parameter as the local parameter. (“In this paper, we revisit the Bayesian CMA-ES formulation and show that there exists indeed an infinity of conjugate prior given by the convex combination of a normal Wishart and normal inverse Wishart Gaussian prior. We first prove that normal Wishart and normal inverse Wishart Gaussian priors have the same update equations except for the mean of the covariance matrix… We then introduce a new prior given by a mixture of normal Wishart and normal inverse Wishart Gaussian prior ” [pg. 2, bottom left col – top right col]) It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Zhang’s/Chen’s teachings in order to implement the Normal-Inverse Wishart model as taught by Benhamou. One would have been motivated to make this modification in order to find the optimal mean and covariance parameter of the multi-variate Normal. [Benhamou, Conclusion, pg. 5]) Regarding claim 16 , it is substantially similar to claim 5 respectively, and is rejected in the same manner, the same art, and reasoning applying. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHAEL H HOANG whose telephone number is (571)272-8491. The examiner can normally be reached Mon-Fri 8:30AM-4:30PM. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kakali Chaki can be reached at (571) 272-3719. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of 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. /MICHAEL H HOANG/PRIMARY EXAMINER, Art Unit 2122 Application/Control Number: 18/512,195 Page 2 Art Unit: 2122 Application/Control Number: 18/512,195 Page 3 Art Unit: 2122 Application/Control Number: 18/512,195 Page 4 Art Unit: 2122 Application/Control Number: 18/512,195 Page 5 Art Unit: 2122 Application/Control Number: 18/512,195 Page 6 Art Unit: 2122 Application/Control Number: 18/512,195 Page 7 Art Unit: 2122 Application/Control Number: 18/512,195 Page 8 Art Unit: 2122 Application/Control Number: 18/512,195 Page 9 Art Unit: 2122 Application/Control Number: 18/512,195 Page 10 Art Unit: 2122 Application/Control Number: 18/512,195 Page 11 Art Unit: 2122 Application/Control Number: 18/512,195 Page 12 Art Unit: 2122 Application/Control Number: 18/512,195 Page 13 Art Unit: 2122 Application/Control Number: 18/512,195 Page 14 Art Unit: 2122 Application/Control Number: 18/512,195 Page 15 Art Unit: 2122 Application/Control Number: 18/512,195 Page 16 Art Unit: 2122 Application/Control Number: 18/512,195 Page 17 Art Unit: 2122 Application/Control Number: 18/512,195 Page 18 Art Unit: 2122 Application/Control Number: 18/512,195 Page 19 Art Unit: 2122 Application/Control Number: 18/512,195 Page 20 Art Unit: 2122 Application/Control Number: 18/512,195 Page 21 Art Unit: 2122 Application/Control Number: 18/512,195 Page 22 Art Unit: 2122 Application/Control Number: 18/512,195 Page 23 Art Unit: 2122 Application/Control Number: 18/512,195 Page 24 Art Unit: 2122
