DETAILED 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 .
This action is responsive to communication filed on 07/27/2026.
Claims 1-11 are pending. Claims 1, 10, and 11 have been amended. Entry of this amendment is accepted and made of record.
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. A subject matter eligibility analysis is set forth below. See MPEP 2106.
Under Step 1 of the analysis, claim 1, belongs to a statutory category namely a method. Likely claims 11, belongs to a statutory category, namely it is a device.
Under Step 2A, prong 1: This part of the eligibility analysis evaluates whether the claim recites a judicial exception. As explained in MPEP 2106.04, subsection II, a claim “recites” a judicial exception when the judicial exception is “set forth” or “described” in the claim.
Claim(s) 1 recite(s) concepts related to mathematical algorithms/concepts, and mental processes and concepts performed in the human mind e.g. observation, evaluation, judgment, opinion of “performing, by a computer, a training phase to generate an individually optimized neural network (NN1, NN2, ...NNk) for said site, said training phase including: training on a training set corresponding to the site associated with said neural network (NN1, NN2, ..., NNk) and based on an extended set of parameters relating to said site, and according to a loss function; a first optimization phase consisting of selecting the optimized set of parameters from said extended set of parameters based upon influence of the parameters of said extended set of parameters on the training of said neural network (NN1, NN2, ..., NNk); a second optimization phase consisting of iteratively modifying at least one structural parameter of said neural network (NN1, NN2, ..., NNk) in order to select at least one optimal structural parameter, based upon influence of said at least one structural parameter on the training of said neural network (NN1, NN2, ..., NNk); acquiring an optimized set (P1, P2, ... Pk) of parameters relating to said site; providing current parameters of said from the optimized set of parameters to a-the generated individually optimized neural network (NN1, NN2, ..., NNk) associated with said site in order to obtain an estimated production (E1, E2,..., Ek);and said method further comprising determining the total estimation (ET) based upon the productions estimated (E1, E2,..., Ek) by each of the generated individually optimized neural networks.”.
Claim(s) 11 recite(s) concepts related to mathematical algorithms/concepts, and mental processes and concepts performed in the human mind e.g. observation, evaluation, judgment, opinion of “perform a training phase to generate an individually optimized neural network (NN1, NN2, ...NNk) for said site, said training phase including: training on a training set corresponding to the site associated with said neural network (NN1, NN2, ..., NNk) and based on an extended set of parameters relating to said site, and according to a loss function; a first optimization phase consisting of selecting the optimized set of parameters from said extended set of parameters based upon influence of the parameters of said extended set of parameters on the training of said neural network (NN1, NN2, ..., NNk); a second optimization phase consisting of iteratively modifying at least one structural parameter of said neural network (NN1, NN2, ..., NNk) in order to select at least one optimal structural parameter, based upon influence of said at least one structural parameter on the training of said neural network (NN1, NN2, ..., NNk); acquiring an optimized set (P1, P2, ... Pk) of parameters relating to said site; provide current parameters of said from the optimized set of parameters to the generated individually optimized neural network (NN1, NN2, ..., NNk) associated with said site in order to obtain an estimated production (E1, E2,..., Ek);and determine the total estimation (ET) based upon the productions estimated (E1, E2,..., Ek) by each of the generated individually optimized neural networks.”.
The concepts discussed above in bold can be considered to describe mental processes, namely concepts performed in the human mind or with pen and paper, and/or mathematical concepts, namely a series of calculations leading to one or more numerical results or answers. Although the claim does not spell out any particular equation or formula being used, the lack of specific equations for individual steps merely points out that the claim would monopolize all possible calculations in performing the steps. These steps recited by the claims, therefore amount to a series of mental or mathematical steps, making these limitations amount to an abstract idea.
Step 2A, prong 2 of the eligibility analysis evaluates whether the claim as a whole integrates the recited judicial exception(s) into a practical application of the exception. This evaluation is performed by (a) identifying whether there are any additional elements recited in the claim beyond the judicial exception, and (b) evaluating those additional elements individually and in combination to determine whether the claim as a whole integrates the exception into a practical application.
This judicial exception is not integrated into a practical application because the abstract idea is not performed by using any particular device and because the “computer implemented method”, “a computer”, “device for estimating an electricity production”, “a processor” and “neural network” respectively recited by respective claims 1 and 11, amounts to the recitation of a general purpose computer used to apply the abstract idea and in which the neural network is used as a tool to implement the abstract idea; the recitation of “acquiring an optimized set (P1, P2, . . . Pk) of parameters relating to said site”, recited by claims 1 and 11, is mere gathering recited at high level of generality and generally linking the abstract idea to a field of use (i.e. electricity production) and the results of the algorithm are merely output/stored as part of insignificant post-solution activity and are not used in any particular matter as to integrate the abstract idea in a practical application.
Under Step 2B, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements, as described above with respect to Step 2A Prong 2, merely amount to a general purpose computer “computer”, “device for estimating (ET)” “processor”, and “neural network”, used to apply the abstract idea and mere data gathering/output recited at a high level of generality (i.e. acquiring an optimized set of parameters) and insignificant extra-solution activity that when further analyzed under Step 2B is found to be well-understood, routine and conventional activities as evidenced by MPEP 2106.05(d)(II); and because the data of performing the algorithm must necessarily be “obtained” and the use of a general purpose computer to implement the abstract idea for performing the algorithm does not amount to significantly more than the recitation of the abstract idea itself.
Therefore, claim 1 and 11 are rejected under 35 U.S.C. 101 as directed to an abstract idea without significantly more.
Dependent claims 2-10 merely expand on the abstract idea by appending additional steps to the mathematical algorithm on their respective independent claim 1.
Dependent claims 2-10 merely expand on the abstract idea by reciting additional steps related to mathematical algorithms/concepts, and mental processes and concepts performed in the human mind e.g. observation, evaluation, judgment, opinion and mere characterization of the data acquired (claims 2-8) and applied for performing the abstract idea i.e. “replacing values of at least one parameter.. with other values, selecting said optimized set of parameters…based on differences in the evolution of said loss function” (claim 4); “each parameter… is tested iteratively” (claim 5), “triggering discrete instances…by iteratively modifying a number of layers…in order to select at least one optimal number of layers and then iteratively modifying a number of neurons… in order to select an optimal number of neurons per layer” (claim 6), “between 3 and 10 optimal numbers of layers are selected” (clam 7), “second optimization phase is based on a NEAT algorithm” (claim 8). The dependent claims do not set forth further additional elements that integrate the abstract idea into a practical application or amount to significantly more than the abstract idea itself.
With respect to dependent claims 9-10 specifically, the claims further recite “said production sites belong to an electrical network of a power production company” (claim 9) and “wherein the program is executed by a computer… to implement the steps of method according to claim 1” (claim 10), respectively. Therefore, these elements are insufficient to integrate the recited abstract idea into a practical application or amount to significantly more than the abstract idea since it amounts to the recitation of a general purpose computer used as a tool to implement the abstract idea and is generally linking the abstract idea to a field of use i.e. power production. Therefore, these claims 2-10 are found ineligible for the reasons described for parent claims 1.
Therefore claims 1-11 are rejected under 35 USC 101 as being directed to non-statutory subject matter.
Response to Arguments
Applicant's arguments filed 07/27/2026 have been fully considered but they are not persuasive.
With respect to claims 1-11, applicant argues with respect to 35 USC 101 rejections that as amended claim 1 is not directed to an abstract idea under Step 2A of the Alice/Mayo framework. (see first paragraph on page 11 of the remarks).
Applicant submits that under Step 2A that claim 1 as amended is explicitly directed to a specific, structured method for generating a set of improved technological tools-the individually optimized neural networks. Applicant submits that the generation process is a practical application that produces an improved set of predictive models configured in a specific and unconventional way to solve a technical problem. Applicant further submits that the claimed invention overcomes the technical problem of improving reliability and scalability of the predictions by creating highly specific models (neural networks) for each production site; and that amended claim 1 now positively recites the specific steps that achieve the technical improvement and that it is not an abstract outcome but a concrete improvement to the functioning of computer-based energy prediction systems (see last three paragraphs on page 11 through third paragraph on page 12 of the remarks).
In response, the examiner disagrees and submits that the claimed language do not reflect the alleged improvement to technology or technical field or to the operation of the computer and that the alleged improvements mentioned is generally linking the use of the judicial exception to a particular technological environment or field of use – see MPEP 2106.05(h) and as such is not indicative of a practical application of abstract idea.
As discussed above, the neural networks are used as a tool to implement the judicial exception and as such do not amount to significantly more than the judicial exception. Furthermore, the alleged improvement in which the applicant relies on is part of the abstract idea and not a practical application of the abstract idea nor an improvement to the technology or technological field. Therefore the claims are patent ineligible under Step 2A.
With respect to Step 2B, applicant argues that even if claims were directed to an abstract idea under Step 2A, the claim recite significantly more under Step 2B (see first paragraph on page 11 of the remarks). Applicant submits that as amended claim 1 recites an inventive concept amounting to “significantly more” that the abstract idea itself. Applicant submits that the Office Action incorrectly characterizes the claimed process as merely “well-understood, routine, conventional activities (see penultimate paragraph on page 12 of the remarks).
In response, the examiner disagrees and submits that applicant has misconstrued the 101 rejections of the claims, in particular with respect to Step 2B. On the outstanding office action and as discussed above with respect to Step 2B, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements, as described above with respect to Step 2A Prong 2, merely amount to a general purpose computer (i.e. “computer”, “device for estimating (ET)” “processor”, and “neural network”), used to apply the abstract idea and mere data gathering/output recited at a high level of generality (i.e. acquiring an optimized set of parameters) and insignificant extra-solution activity that when further analyzed under Step 2B is found to be well-understood, routine and conventional activities as evidenced by MPEP 2106.05(d)(II); and because the data of performing the algorithm must necessarily be “obtained” and the use of a general purpose computer to implement the abstract idea for performing the algorithm. And therefore, it does not amount to significantly more than the recitation of the abstract idea itself.
Applicant argues that the process is not simply applying math and that it is a specific, ordered combination of steps that is not routine or pre-existing and submits that the ordered combination of optimizing both the inputs and the model architecture for each individual production site provides a non-conventional technological solution to the prior art's problem. Applicant submits that this process is explicitly tied to the physical world- it operates on data from physical power production sites (e.g., solar or wind farms, [0004]) to improve the management of a physical electrical network ([0031]). This is far from a purely abstract concept.” (see last paragraph on page 12 of the remarks through second paragraph on page 13 of the remarks). Applicant further submits that the amended claim 1 thus adds meaningful limitations beyond merely implementing an abstract idea on a generic computer and that the specific two-phase optimization provides the "inventive concept" that transforms the claim into a patent- eligible application (see third paragraph on page 13 of the remarks).
In response, the examiner disagrees and submits that as discussed above the features in which the applicant relies (i.e. “selecting an optimized set of parameters from an extended set for each specific production site”, “iteratively modifying at least one structural parameter of said neural network’s structure… I order to select at least one optimal structural parameter”) as an improvement are part of the abstract idea and that as such the argued limitations recite(s) concepts related to mathematical algorithms, and mental processes and concepts performed in the human mind e.g. observation, evaluation, judgment, opinion (i.e. selection of parameters) and do not amount to significantly more than the judicial exception not integrate the judicial exception into a practical application as to be considered significantly more than the abstract idea itself. Examiner further submits that although the claim operates on data from physical power production sites this is generally linking the abstract idea to a field of use and do not amount to significantly more than the judicial exception.
Applicant submits that “the Office Action's own analysis of the prior art confirms the unconventional nature of the claimed method. The Office Action explicitly states that "the closest prior art of record either in singularly or in combination fails to anticipate or render obvious the limitations" of the claims (p. 13). This finding directly contradicts the assertion that the claimed elements are merely "well-understood, routine, and conventional activities" (p. 7). If the specific, ordered combination of a two-phase optimization process to generate individually-tailored neural networks were truly conventional, it would have been taught or suggested by the prior art. Therefore, the claims recite a specific, inventive application that amounts to "significantly more" than the abstract idea itself. (penultimate paragraph on page 12 of the remarks).
In response, the examiner submits that even if the that the argued limitations were new, these steps discussed above still considered to be an abstract idea without more since these argued features are concepts are part of the abstract idea and are merely concepts related to mathematical algorithms, namely a series of calculations leading to one or more numerical results or answers and mental processes and concepts performed in the human mind e.g. observation, evaluation, judgment, namely concepts performed in the human mind or with pen and paper. The examiner further submits that a claim for a new abstract idea is still an abstract idea, so arguing a claim is new is not persuasive by itself.
Furthermore, the combination of steps do not use the data in an unconventional way as to provide an inventive concept and the abstract idea is not applied in an unconventional way as to amount to significantly more than the abstract idea by integrating the abstract idea into a practical application and therefore it is not patent eligible under 35 USC 101.
Moreover, the questions of whether a particular invention meets the "novelty" requirements of 35 U.S.C. § 102 or the "nonobviousness" requirements of § 103 do not affect the determination of whether the invention falls into a category of subject matter that is eligible for patent protection under § 101. Pp. 450 U. S. 185-191.
Even if these mathematical concepts and mental process steps were new, these steps are still considered to be an abstract idea without more since these argued features are concepts that are merely mathematical calculations and metal processes. The examiner further submits a claim for a new abstract idea is still an abstract idea. So arguing a claim is new is not persuasive by itself.
With respect to independent claim 11 and dependent claims 2-9, applicant have presented similar arguments to those with respect to independent claim 1. In response, the examiner disagrees for similar reasons discussed with respect to independent claim 1 above.
Therefore, for the reasons discussed above, claims 1-11 stand rejected under 35 USC 101 as been directed to non-statutory subject matter.
Applicant arguments with respect to 35 USC 112(b) rejections with respect t amended claims 1-11 filed on 07/27/2026 have been fully considered and in view of the amendments of claims 1 and 11 the rejections have been overcome. Rejections made to claims 1-11 under 35 USC 112(b) have been withdrawn.
Conclusion
The prior art made of record cited and not relied upon is considered pertinent to applicant's disclosure.
Zhao et al. CN115049098A disclose: an electricity-water combined system optimal energy flow stochastic optimization method based on data driving (see abstract, para. n0005), wherein the combined electric-water energy flow model is trained using load samples and control variables (para n0007, n0009-n0011), wherein training data is preprocessed to obtain preprocessed electrical load power, water load power, generator output (n0022) and wherein a loss function of the deep learning neural network is calculated and wherein the training of the deep learning neural network is complete when the loss function reaches its minimum value and no longer changes with the number of training iterations (para. n0027-n0028, n0043-n0046).
Han Shuang et al., 2019 Publication "MId-to-long term wind and photovoltaic power generation prediction based on copula function and long short term memory network", (hereinafter Han)- Han disclose independent mid to long term wind and photovoltaic power generation for power grid plan improvement dispatching optimization, management development and consumption enhancement power generation predictions performed by prediction models (see abstract). Han further disclose the use of LSTM neural network which is established under different inputs to choose best inputs (see section 4.1) and to use root mean square percentage error to evaluate overall error during predicted period (see section 4.2). Han disclose the predicted results are based on the LSTM neural network prediction model which showed better persistence prediction model (see section 4.4).
Jenatton et al. US20230206030A1 disclose a method of training an ensemble comprising K neural networks to perform a machine learning task, wherein K is an integer greater than one, wherein each of the K neural networks comprises a plurality of neural network layers having respective parameters, wherein the plurality of neural network layers includes a first neural network layer that, for each of the K neural networks (see abstract, claim 1).
Arnold US20240078232A1 disclose a database system is operable to generate a query operator execution flow for a query that includes a plurality of operators implementing a plurality of parallelized optimization processes configured to facilitate generating of a machine learning model. The query operator execution flow is executed in conjunction with executing the query based on executing the plurality of operators. Executing each of the plurality of parallelized optimization processes includes generating a corresponding set of candidate model coefficients of a plurality of sets of candidate model coefficients. A corresponding set of candidate model coefficients is based on processing the set of best positions generated via the second type of optimization algorithm. The machine learning model is generated in executing the query based on selection of a most favorable set of candidate model coefficients from a plurality of sets of candidate model coefficients outputted via the plurality of parallelized optimization processes (see abstract).
Nachum et al. US11,315,019 disclose a system for training neural networks, the system comprising a neural network shrinking engine that is configured to: receive a neural network being trained; generate a reduced neural network by a shrinking process comprising: training the neural network based on a shrinking engine loss function that includes terms penalizing active neurons of the neural network, wherein: a neuron is active if dropping the neuron from the neural network, along with all its incoming and outgoing connections, would affect outputs of the neural network generated in response to processing inputs to the neural network; removing inactive neurons from the neural network; a neural network expansion engine that is configured to: receive the neural network being trained; generate an expanded neural network by an expansion process comprising: adding new neurons to the neural network; training the neural network based on an expanding engine loss function; a training subsystem that is configured to, at each of a plurality of time steps: generate a reduced neural network for the time step by providing an expanded neural network for a previous time step to the neural network shrinking engine; and generate an expanded neural network for the time step by providing the reduced neural network for the time step to the neural network expansion engine (col 1, ll. 25-48; col. 3, ll. 33-45). Nachum further disclose a loss function that penalize active neurons of the parameter of a batch normalization layer of the neural network (col. 1, ll. 54-59) and further disclose “In some implementations, training the neural network based on the expanding engine loss function comprises: training only parameters of newly added neurons of the neural network for a first number of training iterations; training parameters of each neuron of the neural network for a second number of training iterations after the first number of training iterations” (col. 3, ll. 18-24).
Asad et al. US 12,260,336 disclose A computer implemented method of training a neural network configured to combine a set of coefficients with respective input data values. So as to train a test implementation of the neural network, sparsity is applied to one or more of the coefficients according to a sparsity parameter, the sparsity parameter indicating a level of sparsity to be applied to the set of coefficients; the test implementation of the neural network is operated on training input data using the coefficients so as to form training output data; in dependence on the training output data, assessing the accuracy of the neural network; the sparsity parameter is updated in dependence on the accuracy of the neural network; and a runtime implementation of the neural network is configured in dependence on the updated sparsity parameter (abstract). Asad further disclose The accuracy of the training output data can be assessed, e.g. using a loss function. The sets of coefficients can be updated in dependence on the accuracy of the training output data through the processes called gradient descent and back-propagation. For example, the sets of coefficients can be updated in dependence on the loss of the training output data determined using the loss function. Back-propagation can be considered to be a process of calculating gradients for each coefficient with respect to a loss function. This can be achieved by using chain rule starting at the final output of the loss function and working backwards to each layer's coefficients. Once all gradients are known, a gradient descent (or its derivative) algorithm can be used to update each coefficient according to its gradients calculated through back-propagation. Gradient descent can be performed in dependence on a learning rate parameter, which indicates the degree to which the coefficients can be changed in dependence on the gradients at each iteration of the training process. These steps can be repeated, so as to iteratively update the sets of coefficients (see col. 14, ll. 1-20).
However the closest prior art of record either in singularly or in combination fails to anticipate or render obvious the limitations of acquiring an optimized set (P1, P2, . . . Pk) of parameters relating to said site; providing parameters of said optimized set of parameters to a neural network (NN1, NN2, . . . , NNk) associated with said site in order to obtain an estimated production (E1, E2, . . . , Ek); said estimation (ET) being determined based upon the productions estimated (E1, E2, . . . Ek) by each of said neural networks (NN1, NN2, . . . , NNk); each of said neural networks (NN1, NN2, . . . , NNk) being previously determined during a training phase including: training on a training set corresponding to the site associated with said neural network (NN1, NN2, . . . , NNk) and based on an extended set of parameters relating to said site, and according to a loss function; a first optimization phase consisting of selecting the optimized set of parameters from said extended set of parameters based upon influence of the parameters of said extended set of parameters on the training of said neural network (NN1, NN2, . . . NNk); a second optimization phase consisting of iteratively modifying at least one structural parameter of said neural network (NN1, NN2, . . . , NNk) in order to select at least one optimal structural parameter, based upon influence of said at least one structural parameter on the training of said neural network (NN1, NN2, . . . , NNk), without the use of impermissible hindsight.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to YARITZA H PEREZ BERMUDEZ whose telephone number is (571)270-1520. The examiner can normally be reached Monday-Friday.
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, Shelby A Turner can be reached at (571) 272-6334. 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.
/YARITZA H PEREZ BERMUDEZ/Examiner, Art Unit 2857
/JORDAN L JACKSON/Primary Examiner, Art Unit 2857
