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 .
Detailed Action
The following action is in response to the communication(s) received on 03/16/2026.
As of the claims filed 03/16/2026:
Claims 1, 10, 13, and 22 have been amended.
Claim 6 has been withdrawn.
Claims 1-5 and 7-22 are now pending.
Claims 1, 10, and 22 are independent claims.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 01/30/2026 were filed in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Response to Arguments
Applicant’s arguments filed 03/26/2026 have been fully considered, but are not fully persuasive.
Regarding the patent eligibility rejection under 35 U.S.C. 101:
Applicant asserts that the present claimed invention provides an improvement to prediction quality and uncertainty quantification in a computationally efficient manner (p.10 ¶2). Examiner respectfully submits that, as currently recited, the claims are directed to selecting top performing neural networks out of an ensemble of candidate trained neural networks. The amended limitation merely adds a detail involved in generating a list of candidate neural networks. Thus, as currently recited, “selecting K neural networks” is merely an abstract idea aided by a computer.
Applicant further asserts that Claim 10 has been amended to clarify the technical improvement stated above (p.10 last ¶). As described above, the improvements recited in the claimed invention remain directed to an abstract idea.
Regarding the art rejection under 35 U.S.C. 103:
The amendments have overcome the art rejections for claims 1-9 and 18-22; thus, the rejections for these claims have been withdrawn. However, the art rejection remains for claims 10-17.
Applicant asserts that the amendments to claim 10 overcomes the art rejection as Levesque teaches a greedy selection of the model when updating the ensemble (p.14 last ¶). Examiner respectfully submits that the claims do not suggest against the greedy selection method from Levesque. In other words, the iterative selection of the final model is still a selection of the final model within the set of candidate trained neural networks of N different hyperparameters. Thus, Levesque remains teaching claims 10-17.
Applicant further asserts that Alhamdoosh does not cure the deficiencies of Levesque. (p.15 ¶4). Examiner respectfully submits that the argument is moot as Levesque remains teaching these limitations.
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 10-17 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claim 10 recites method, thus a process, one of the four statutory categories of patentable subject matter (Step 1). However, Claim 10 further recites:
identifying a set of N different hyperparameters for training a neural network having parameters to perform the machine learning task, wherein N is an integer greater than one, which is an evaluation or judgement that can be performed in the human mind;
generating a set of first candidate trained neural networks that comprises a plurality of trained neural networks for each of the N different hyperparameters by, for each of the N different hyperparameters, which is an evaluation or judgement that can be performed in the human mind;
selecting a plurality of different initializations, which is an evaluation or judgement that can be performed in the human mind;
generating the ensemble of K neural networks by selecting K neural networks from the first candidate trained neural networks, which is an evaluation or judgement that can be performed in the human mind.
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2, the claim does not include any additional elements which integrate the abstract idea into a practical application, since the additional elements consist of:
for values of the parameters of the neural network, which merely specifies the particular field of use or particular technological environment in which the abstract idea is to be performed, which by MPEP 2106.05(h) cannot integrate the abstract idea into a practical application;
and for each of the different initializations, training a corresponding neural network with (i) the different hyperparameters and (ii) parameter values initialized using the different initialization to generate a trained neural network, as the performance of an abstract idea on a computer is not more than instructions to "apply it" on a computer, which by MPEP 2106.05(f) cannot integrate an abstract idea into a practical application.
Thus, the claim is directed towards an abstract idea.
Further, the additional elements, alone or in combination, do not provide significantly more than the abstract idea itself, because the particular field of use or particular technological environment (MPEP 2106.05(h)), implementation on a computer (MPEP 2106.05(f)), cannot provide significantly more, and the combination of additional elements does not provide an inventive concept. Thus, the claim is ineligible.
Claim 11, dependent upon Claim 10, further recites
applying a hyperparameter search technique to identify M best-performing hyperparameters for the machine learning task, wherein M is an integer that is greater than N, which is an evaluation or judgement that can be performed in the human mind;
and selecting, from the M best-performing hyperparameters, N hyperparameters using a first ensemble selection technique, which is an evaluation or judgement that can be performed in the human mind.
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2 and 2B, the claim does not recite any new additional elements which could integrate the abstract idea into a practical application or provide significantly more than the abstract idea itself. Thus, the claim is ineligible.
Claim 12, dependent upon Claim 11, further recites
N is equal to K, which is merely a detail of an abstract idea (selecting… N hyperparameters).
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2 and 2B, the claim does not recite any new additional elements which could integrate the abstract idea into a practical application or provide significantly more than the abstract idea itself. Thus, the claim is ineligible.
Claim 14, dependent upon Claim 11, further recites
generating a set of M second candidate neural networks..., which is an evaluation or judgement that can be performed in the human mind;
generating, from the M second candidate neural networks, a first ensemble of N candidate neural networks by repeatedly adding to the first ensemble by selecting from the M candidate neural networks the candidate neural network that, if added to the first ensemble, would result in a largest increase in performance of the first ensemble, which is an evaluation or judgement that can be performed in the human mind;
and selecting, as the N hyperparameters, the hyperparameters..., which is an evaluation or judgement that can be performed in the human mind.
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2, the claim does not include any additional elements which integrate the abstract idea into a practical application, since the additional elements consist of:
that have each been trained using a different one of the M best-performing hyperparameters, as the performance of an abstract idea on a computer is not more than instructions to "apply it" on a computer, which by MPEP 2106.05(f) cannot integrate an abstract idea into a practical application;
used to train the N candidate neural networks in the first ensemble, which merely specifies the particular field of use or particular technological environment in which the abstract idea is to be performed, which by MPEP 2106.05(h) cannot integrate the abstract idea into a practical application.
Thus, the claim is directed towards an abstract idea.
Further, the additional elements, alone or in combination, do not provide significantly more than the abstract idea itself, because the particular field of use or particular technological environment (MPEP 2106.05(h)), implementation on a computer (MPEP 2106.05(f)), cannot provide significantly more, and the combination of additional elements does not provide an inventive concept. Thus, the claim is ineligible.
Claim 15, dependent upon Claim 10, further recites
generating the ensemble using a second ensemble generation technique, which is an evaluation or judgement that can be performed in the human mind.
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2 and 2B, the claim does not recite any new additional elements which could integrate the abstract idea into a practical application or provide significantly more than the abstract idea itself. Thus, the claim is ineligible.
Claim 16, dependent upon Claim 15, further recites
generating... the ensemble of K neural networks by repeatedly adding a first candidate trained neural network to the ensemble by selecting from the set of first candidate trained neural networks, the first candidate trained neural network that, if added to the ensemble, would result in a largest increase in performance of the ensemble, which is an evaluation or judgement that can be performed in the human mind.
Thus, the claim recites an abstract idea under Step 2A Prong 1.
Under Step 2A Prong 2, the claim does not include any additional elements which integrate the abstract idea into a practical application, since the additional elements consist of:
from the set of first candidate trained neural networks, which merely specifies the particular field of use or particular technological environment in which the abstract idea is to be performed, which by MPEP 2106.05(h) cannot integrate the abstract idea into a practical application.
Thus, the claim is directed towards an abstract idea.
Further, under Step 2B, the additional element does not provide significantly more than the abstract idea itself, because the particular field of use or particular technological environment (MPEP 2106.05(h)) cannot provide significantly more. Thus, the claim is ineligible.
Claim 17, dependent upon Claim 10, further recites no additional abstract ideas. However:
Under Step 2A Prong 2, the claim does not include any additional elements which integrate the abstract idea into a practical application, since the additional elements consist of:
the different initializations for the parameters of the neural network are different random initializations of values of the parameters of the neural network, which merely specifies the particular field of use or particular technological environment in which the abstract idea is to be performed, which by MPEP 2106.05(h) cannot integrate the abstract idea into a practical application.
Thus, the claim is directed towards an abstract idea.
Further, under Step 2B, the additional element does not provide significantly more than the abstract idea itself, because the particular field of use or particular technological environment (MPEP 2106.05(h)) cannot provide significantly more. Thus, the claim is ineligible.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 10-17 are rejected under 35 U.S.C. 103 as being unpatentable over Levesque in view of Alhamdoosh et al., "Fast decorrelated neural network ensembles with random weights" (hereinafter Alhamdoosh).
Regarding Claim 10, Levesque teaches:
A method of training an ensemble comprising K neural networks to perform a machine learning task, wherein K is an integer greater than one, and wherein the method comprises: (Levesque [Algorithm 1, H]
PNG
media_image1.png
338
418
media_image1.png
Greyscale
) (Note: H corresponds to the K neural networks)
identifying a set of N different hyperparameters for training a neural network having parameters to perform the machine learning task, wherein N is an integer greater than one; (Levesque [p.2 left ¶3] The behavior of a learning algorithm A is often tunable with regards to a set of external parameters, called hyperparameters γ = {γ1, γ2, . . . } ∈ Γ, which are not learned during training…
[Alg. 1, Γ; i in line 2; line 7]
PNG
media_image2.png
338
418
media_image2.png
Greyscale
)
generating a set of first candidate trained neural networks that comprises a plurality of trained neural networks for each of the N different parameters by, for each of the N different hyperparameters: (Levesque [Algorithm 1, line 6]
PNG
media_image1.png
338
418
media_image1.png
Greyscale
) (Note: f(γ|E) at each loop corresponds to each of the set of trained first candidate neural networks)
Levesque does not teach, but Alhamdoosh further teaches:
selecting a plurality of different initializations for values of the parameters of the neural network; and (Alhamdoosh, [p.2 ¶4] Our proposed algorithm (termed as DNNE) randomly initializes the hidden layer parameters of base RVFL networks, and then employs the least square method with negative correlation learning scheme to analytically calculate the output weights of these base networks)
Alhamdoosh and Levesque are analogous to the present invention because both are from the same field of endeavor of ensemble neural networks. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to implement the random initialization of the parameters of each model of the ensemble from Alhamdoosh into Levesque’s hyperparameter ensemble network. The motivation would be to “to develop a fast solution on building neural network ensembles” (Alhamdoosh, p.2 ¶4).
Levesque, via Levesque/Alhamdoosh, further teaches:
for each of the different initializations, training a corresponding neural network with (i) the different hyperparameters and (ii) parameter values initialized using the different initialization to generate a trained neural network; and (Levesque [Alg. 1, line 8] A corresponds to the training of the corresponding neural network)
generating the ensemble of K neural networks by selecting K neural networks from the first candidate trained neural networks. (Levesque [Alg. 1, line 10-12 “E”])
Regarding Claim 11, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 10. Levesque, via Levesque/Alhamdoosh, further teaches:
The method of claim 10, wherein identifying a set of N different hyperparameters for training a neural network in the ensemble comprises:
applying a hyperparameter search technique to identify M best-performing hyperparameters for the machine learning task, wherein M is an integer that is greater than N; and
selecting, from the M best-performing hyperparameters, N hyperparameters using a first ensemble selection technique. (Levesque
PNG
media_image3.png
422
407
media_image3.png
Greyscale
) (Note: each cross corresponds to the M best-performing hyperparameters; each circle E from the result of the ensemble optimization corresponds to the selected N hyperparameters using the ensemble selection technique)
Regarding Claim 12, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 11. Levesque, via Levesque/Alhamdoosh, further teaches:
The method of claim 11, wherein N is equal to K. (Levesque [Fig.1, E; Algorithm 1, E]
PNG
media_image1.png
338
418
media_image1.png
Greyscale
) (Note: E corresponds to the K neural networks)
Regarding Claim 13, Wen/Levesque respectively teaches and incorporates the claimed limitations and rejections of Claim 1. Levesque, via Wen/Levesque, further teaches:
The method of claim 1, wherein the hyperparameter search technique is random search. (Levesque [p.6 right ¶2] In every case, the prior on the objective function is a Gaussian Process (GP) with Matern-52 kernel using automatic relevance determination. The noise, amplitude, and length-scale parameters are obtained through slice sampling... The slice sampling of GP hyperparameters for ensemble optimization must be reinitialized at every iteration given that different ensembles E can change the properties of the optimized function drastically.) (Note: sampling corresponds to randomly selecting; thus, the slice sampling corresponds to the random hyperparameter search technique)
Regarding Claim 14, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 11. Levesque, via Levesque/Alhamdoosh, further teaches:
The method of claim 11, wherein selecting the N hyperparameters using the ensemble selection technique comprises:
generating a set of M second candidate neural networks that have each been trained using a different one of the M best-performing hyperparameters; (Levesque [p.3 left ¶1] At each iteration, given a pool of trained classifiers H to select from…) (Note: H corresponds to M second candidate neural networks.)
generating, from the M second candidate neural networks, a first ensemble of N candidate neural networks by repeatedly adding to the first ensemble by selecting from the M candidate neural networks the candidate neural network that, if added to the first ensemble, would result in a largest increase in performance of the first ensemble; and
selecting, as the N hyperparameters, the hyperparameters used to train the N candidate neural networks in the first ensemble. (Levesque [p.3 left ¶1]… a new classifier is added to the ensemble, selected according to the minimum ensemble generalization error. At the first iteration, the classifier added is simply the single best classifier. At step t, given the ensemble E = {he1 , he2 , . . . , het−1 }, the next classifier is chosen to minimize the empirical error on the validation dataset when added to E:
PNG
media_image4.png
96
388
media_image4.png
Greyscale
where g(xi , E) is a function combining the predictions of the classifiers in E on sample xi . In this case, the combination rule is majority voting, as it is less prone to overfitting.) (Note: the subsequent classifier chosen to minimize the empirical error corresponds to generating the first ensemble; each h corresponds to each of the N hyperparameters)
Regarding Claim 15, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 10. Levesque, via Levesque/Alhamdoosh, further teaches:
The method of claim 10, wherein generating the ensemble of K neural networks comprises generating the ensemble using a second ensemble generation technique. (Levesque [p.3 left ¶1] At each iteration, given a pool of trained classifiers H to select from, a new classifier is added to the ensemble, selected according to the minimum ensemble generalization error. At the first iteration, the classifier added is simply the single best classifier. At step t, given the ensemble E = {he1 , he2 , . . . , het−1 }, the next classifier is chosen to minimize the empirical error on the validation dataset when added to E:
PNG
media_image4.png
96
388
media_image4.png
Greyscale
where g(xi , E) is a function combining the predictions of the classifiers in E on sample xi . In this case, the combination rule is majority voting, as it is less prone to overfitting.) (Note: the subsequent classifier chosen to minimize the empirical error corresponds to generating the first ensemble; each h corresponds to each of the N hyperparameters)
Regarding Claim 16, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 15. Levesque, via Levesque/Alhamdoosh, further teaches:
The method of claim 15, wherein generating the ensemble of K neural networks comprises:
generating, from the set of first candidate trained neural networks, the ensemble of K neural networks by repeatedly adding a first candidate trained neural network to the ensemble by selecting from the set of first candidate trained neural networks, the first candidate trained neural network that, if added to the ensemble, would result in a largest increase in performance of the ensemble. (Levesque [p.3 left ¶1] At each iteration, given a pool of trained classifiers H to select from, a new classifier is added to the ensemble, selected according to the minimum ensemble generalization error. At the first iteration, the classifier added is simply the single best classifier. At step t, given the ensemble E = {he1 , he2 , . . . , het−1 }, the next classifier is chosen to minimize the empirical error on the validation dataset when added to E:
PNG
media_image4.png
96
388
media_image4.png
Greyscale
where g(xi , E) is a function combining the predictions of the classifiers in E on sample xi . In this case, the combination rule is majority voting, as it is less prone to overfitting.) (Note: the subsequent classifier chosen to minimize the empirical error corresponds to generating the first ensemble; each h corresponds to each of the N hyperparameters)
Regarding Claim 17, Levesque/Alhamdoosh respectively teaches and incorporates the claimed limitations and rejections of Claim 10. Alhamdoosh, via Levesque/Alhamdoosh, further teaches:
The method of claim 10, wherein the different initializations for the parameters of the neural network are different random initializations of values of the parameters of the neural network. (Alhamdoosh, [p.2 ¶4] Our proposed algorithm (termed as DNNE) randomly initializes the hidden layer parameters of base RVFL networks, and then employs the least square method with negative correlation learning scheme to analytically calculate the output weights of these base networks)
Allowable Subject Matter
Claims 1-9 and 18-22 are allowed.
Conclusion
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 JOSEP HAN whose telephone number is (703)756-1346. The examiner can normally be reached Mon-Fri 9am-5pm.
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 on (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.
/J.H./Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights