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 is a Final Office Action in response to the RCE on application 18/375,098 entitled "METHOD AND SYSTEM FOR PROVIDING SYNTHETIC NEURAL DATA MODELS" filed on September 30, 2025, with claims 1, 2, 4-6, 8-11, 13-15 and 17-26 pending.
Status of Claims
Claims 1, 10, 19, 23 and 24 have been amended and are hereby entered.
Claims 3, 7, 12 and 16 were previously cancelled.
Claims 1, 2, 4-6, 8-11, 13-15 and 17-26 are pending and have been examined.
Response to Amendment
The amendment filed March 25, 2026, has been entered. Claims 1, 2, 4-6, 8-11, 13-15 and 17-26 remain pending in the application. Applicant’s amendments to the Specification, Drawings, and/or Claims have been noted in response to the Non-Final Office Action mailed December 29, 2025.
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, 2, 4-6, 8-11, 13-15 and 17-26 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Please see MPEP 2106 for additional information regarding Patent Subject Matter Eligibility Guidance.
Claims 1, 2, 4-6, 8-11, 13-15 and 17-26 are directed to a method/process, machine/apparatus, (article of) manufacture, or composition of matter, which are/is one of the statutory categories of invention, which are/is one of the statutory categories of invention. (Step 1: YES).
The claimed invention is directed to an abstract idea without significantly more.
Independent Claim 1 recites:
“A method for providing a synthetic … data model, the method being implemented …the method comprising:
generating, …, a model that is configured to simulate …;
appending, … at least one agent to the model, the at least one agent relating to a software component that sends at least one request to the model based on a predetermined timestep;
assigning… a fixed grid to the model, the fixed grid including a tick size that corresponds to a fixed granularity in which consecutive values are spaced apart by a distance that corresponds to the tick size;
calibrating, … each of the at least one agent by using a calibration data set;
obtaining, …, historical data;
inputting, … the model and the historical data …and
beginning, …, …extension of the model and using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss;
identifying, …, a multivariate Gaussian distribution mixture that characterizes…;
assessing, ..a validation loss parameter of the model and the neural network extension:
plotting …extension learning curve for the validation loss parameter;
and stopping, …, extension when the validation loss parameter begins to increase.”
These limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. Specific instances include instructions for “obtaining…historical (order book) data” and “tick size” recites a fundamental economic principles or practice and/or commercial or legal interactions. Furthermore, the Specification reads:
[0072] the electronic communication network may relate to a computerized system that automatically matches buy and sell orders for securities in a financial market.
[0007] the at least one model may include support for a plurality of orders that relate to at least one from among a market order, a limit order, and a cancel order, the support may enable the plurality of orders on a bid and ask side.
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a fundamental economic, commercial, or financial action, principle, or practice then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas.
Additionally, these limitations, under their broadest reasonable interpretation, cover performance of the limitation as Mathematical Concepts. Specific instances include instructions that "providing a synthetic neural data model" and "a model that is configured to simulate" and "assigning...a fixed grid to the model" and "calibrating...using a calibration data set" and "using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss " and “a multivariate Gaussian distribution mixture” recite Mathematical Concepts. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as mathematical relationships or mathematical calculations then it falls within the “Mathematical Concepts” grouping of abstract ideas.
Accordingly, the claim recites an abstract idea. (Step 2A-Prong 1: YES. The claims recite an abstract idea).
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of:
[by at least one processor][by the at least one processor][ an electronic communication network][ the electronic communication network]:
merely applying computer processing, storage, and networking technology as tools to perform an abstract idea
[to a neural network][training][via the neural network][neural network] [training of a neural network ][training of the neural network ][via the neural network][training ][a neural network ] :
merely applying the generic machine learning functionality to perform the abstract idea.
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For example, the Applicant’s Specification reads:
[0035] the computer system 102 may include, or be included within, any one or more computers, servers, systems, communication networks or cloud environment
[0036] The computer system 102, or portions thereof, may be implemented as, or incorporated into, various devices, such as a personal computer
[0042] Furthermore, the computer system 102 may include any additional devices, components, parts, peripherals, hardware, software, or any combination thereof which are commonly known and understood
[0090] In another exemplary embodiment, consistent with present disclosures, the neural network extension may be trained by using an optimizer such as, for example, an ADAM optimizer with learning rate 10-3 up to epoch 110, then 10-4, and a minibatch size of 100. The network weights may be initialized by using an initialization function such as, for example, a XAVIER initialization that is associated with the uniform distribution.
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, Claim 1 is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, the additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. The claim further defines the abstract idea and hence is abstract for the reasons presented above. The claim does not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Therefore, the claim is directed to an abstract idea. Thus, the claim is not patent eligible. (Step 2B: NO. The claim does not provide significantly more)
Dependent Claims recite additional elements.
This judicial exception is not integrated into a practical application. In particular, the recited additional elements of
Claim 2: (none found: does not include additional elements and merely narrows the abstract idea)
Claim 4: (none found: does not include additional elements and merely narrows the abstract idea)
Claim 5:
“by the at least one processor”: merely applying computer processing technologies as a tool to perform an abstract idea
“training the neural network”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 6:
“by the at least one processor”: merely applying computer processing technologies as a tool to perform an abstract idea
“neural network”, “hidden layer”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 8:
“by the at least one processor”, “electronic communication network”: merely applying computer processing, networking, and display technologies as a tool to perform an abstract idea
“transmitting”: insignificant extra-solution activity to the judicial exception of data gathering
Claim 9:
“machine learning”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 21: (none found: does not include additional elements and merely narrows the abstract idea)
Claim 23:
“training”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 25:
“electronic communication network”: merely applying networking technologies as a tool to perform an abstract idea
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For example, the Applicant’s Specification reads:
[0035] the computer system 102 may include, or be included within, any one or more computers, servers, systems, communication networks or cloud environment
[0036] The computer system 102, or portions thereof, may be implemented as, or incorporated into, various devices, such as a personal computer
[0042] Furthermore, the computer system 102 may include any additional devices, components, parts, peripherals, hardware, software, or any combination thereof which are commonly known and understood
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, the claim is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. Dependent claims further define the abstract idea that is present in their respective independent claims and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination.
Transmitting a message is not an abstract idea, it is post solution activity. Under step 2B support that transmitting the message is found in the MPEP 2106.05 cases OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network) and Ameranth, 842 F.3d at 1241-42, 120 USPQ2d at 1854-55.
Therefore, the dependent claims are directed to an abstract idea. Thus, the dependent claims are not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Independent Claim 10 recites:
“A … configured to implement an execution of a method for providing a synthetic … data model, …:
generate a model that is configured to simulate …;
append, … at least one agent to the model, the at least one agent relating to a software component that sends at least one request to the model based on a predetermined timestep;
assign… a fixed grid to the model, the fixed grid including a tick size that corresponds to a fixed granularity in which consecutive values are spaced apart by a distance that corresponds to the tick size;
calibrate, … each of the at least one agent by using a calibration data set;
obtain historical data;
input, … the model and the historical data …
begin…, a neural network extension of the model and using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss;
identify, …a multivariate Gaussian distribution mixture that characterizes …
assess a validation loss parameter of the model …extension;
plot a … extension learning curve for the validation loss parameter;
and stop … extension when the validation loss parameter begins to increase.”
These limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. Specific instances include instructions to “obtain historical (order book) data” recites a fundamental economic principles or practice and/or commercial or legal interactions. Furthermore, the Specification reads:
[0072] the electronic communication network may relate to a computerized system that automatically matches buy and sell orders for securities in a financial market.
[0007] the at least one model may include support for a plurality of orders that relate to at least one from among a market order, a limit order, and a cancel order, the support may enable the plurality of orders on a bid and ask side.
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a fundamental economic, commercial, or financial action, principle, or practice then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas.
Additionally, these limitations, under their broadest reasonable interpretation, cover performance of the limitation as Mathematical Concepts. Specific instances include instructions that "providing a synthetic neural data model" and "a model that is configured to simulate" and "assigning...a fixed grid to the model" and "calibrating...using a calibration data set" and "using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss " and “a multivariate Gaussian distribution mixture” recite Mathematical Concepts. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as mathematical relationships or mathematical calculations then it falls within the “Mathematical Concepts” grouping of abstract ideas.
Accordingly, the claim recites an abstract idea. (Step 2A-Prong 1: YES. The claims recite an abstract idea).
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of:
[computing device comprising: a processor; a memory; and a communication interface coupled to each of the processor and the memory, wherein the processor is configured] [computing device] [by at least one processor][by the at least one processor]:
merely applying computer processing, storage, and networking technology as tools to perform an abstract idea
[to a neural network][training][via the neural network][into a neural network] :
merely applying the generic machine learning functionality to perform the abstract idea.
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For support from the Applicant’s Specification, see the analysis as applied to Independent Claim 1 earlier. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, Claim 10 is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, the additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. The claim further defines the abstract idea and hence is abstract for the reasons presented above. The claim does not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Therefore, the claim is directed to an abstract idea. Thus, the claim is not patent eligible. (Step 2B: NO. The claim does not provide significantly more)
Dependent Claims recite additional elements.
This judicial exception is not integrated into a practical application. In particular, the recited additional elements of
Claim 11:
“computing device”: merely applying computer processing technologies as a tool to perform an abstract idea
Claim 14:
“computing device”, “processor”: merely applying computer processing technologies as a tool to perform an abstract idea
“train the neural network”, “neural network”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 15:
“computing device”, “processor”: merely applying computer processing technologies as a tool to perform an abstract idea
“neural network”, “hidden layer”: merely applying machine learning technology as a tool to perform an abstract idea
Claim 17:
“computing device”, “processor”, “electronic communication network”: merely applying computer processing, networking, and display technologies as a tool to perform an abstract idea
“transmitting”: insignificant extra-solution activity to the judicial exception of data gathering
Claim 18:
“computing device”: merely applying computer processing technologies as a tool to perform an abstract idea
“machine learning”: merely applying machine learning technology as a tool to perform an abstract idea
Claims 20 and 22:
“processor”: merely applying computer processing technologies as a tool to perform an abstract idea
Claim 24:
“processor”: merely applying computer processing technologies as a tool to perform an abstract idea
Claim 25:
“electronic communication network”: merely applying networking technologies as a tool to perform an abstract idea
Claim 26:
“processor”: merely applying computer processing technologies as a tool to perform an abstract idea
“electronic communication network”: merely applying networking technologies as a tool to perform an abstract idea
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For example, the Applicant’s Specification reads:
[0035] the computer system 102 may include, or be included within, any one or more computers, servers, systems, communication networks or cloud environment
[0036] The computer system 102, or portions thereof, may be implemented as, or incorporated into, various devices, such as a personal computer
[0042] Furthermore, the computer system 102 may include any additional devices, components, parts, peripherals, hardware, software, or any combination thereof which are commonly known and understood
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, the claim is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. Dependent claims further define the abstract idea that is present in their respective independent claims and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination.
Transmitting a message is not an abstract idea, it is post solution activity. Under step 2B support that transmitting the message is found in the MPEP 2106.05 cases OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network) and Ameranth, 842 F.3d at 1241-42, 120 USPQ2d at 1854-55.
Therefore, the dependent claims are directed to an abstract idea. Thus, the dependent claims are not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Independent Claim 19 recites:
“A …for providing a synthetic … data model, …to:
generate a model that is configured to simulate …;
append, … at least one agent to the model, the at least one agent relating to a software component that sends at least one request to the model based on a predetermined timestep;
assign… a fixed grid to the model, the fixed grid including a tick size that corresponds to a fixed granularity in which consecutive values are spaced apart by a distance that corresponds to the tick size;
calibrate, … each of the at least one agent by using a calibration data set;
obtain historical data;
input, … the model and the historical data …
begin…, a neural network extension of the model and using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate training loss;
identify, … a multivariate Gaussian distribution mixture that characterizes …;
assess a validation loss parameter of the model and … extension;
plot a … extension learning curve for the validation loss parameter; and
stop …extension when the validation loss parameter begins to increase”
These limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. Specific instances include instructions to “obtain historical (order) book data” recites a fundamental economic principles or practice and/or commercial or legal interactions. Furthermore, the Specification reads:
[0072] the electronic communication network may relate to a computerized system that automatically matches buy and sell orders for securities in a financial market.
[0007] the at least one model may include support for a plurality of orders that relate to at least one from among a market order, a limit order, and a cancel order, the support may enable the plurality of orders on a bid and ask side.
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a fundamental economic, commercial, or financial action, principle, or practice then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas.
Additionally, these limitations, under their broadest reasonable interpretation, cover performance of the limitation as Mathematical Concepts. Specific instances include instructions that "providing a synthetic neural data model" and "a model that is configured to simulate" and "assigning...a fixed grid to the model" and "calibrating...using a calibration data set" and "using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss " and “a multivariate Gaussian distribution mixture” recite Mathematical Concepts. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as mathematical relationships or mathematical calculations then it falls within the “Mathematical Concepts” grouping of abstract ideas.
Accordingly, the claim recites an abstract idea. (Step 2A-Prong 1: YES. The claims recite an abstract idea).
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of:
[non-transitory computer readable storage medium storing instructions] [the storage medium comprising executable code which, when executed by a processor, causes the processor] [an electronic communication network]:
merely applying computer processing, storage, and networking technology as tools to perform an abstract idea
[to a neural network][training][via the neural network][neural network] :
merely applying the generic machine learning functionality to perform the abstract idea.
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For support from the Applicant’s Specification, see the analysis as applied to Independent Claim 1 earlier. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, Claim 19 is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, the additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. The claim further defines the abstract idea and hence is abstract for the reasons presented above. The claim does not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Therefore, the claim is directed to an abstract idea. Thus, the claim is not patent eligible. (Step 2B: NO. The claim does not provide significantly more)
Dependent Claim recites additional elements.
This judicial exception is not integrated into a practical application. In particular, the recited additional elements of
Claim 20:
“storage medium”: merely applying computer storage technologies as a tool to perform an abstract idea
are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer components and/or electronic processes. For support from the Applicant’s Specification, see the analysis as applied to Independent Claim 1 earlier. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea and are at a high level of generality. Therefore, the claim is directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The additional elements merely add instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, see MPEP 2106.05(f). Accordingly, these additional elements, do not change the outcome of the analysis, when considered separately and as an ordered combination. Dependent claims further define the abstract idea that is present in their respective independent claims and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination.
Transmitting a message is not an abstract idea, it is post solution activity. Under step 2B support that transmitting the message is found in the MPEP 2106.05 cases OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network) and Ameranth, 842 F.3d at 1241-42, 120 USPQ2d at 1854-55.
Therefore, the dependent claim is directed to an abstract idea. Thus, the dependent claim is not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis 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.
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.
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 set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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.
Claims 1, 5, 6, 9, 10, 14, 15, 18, 19, and 21-24 are rejected under 35 U.S.C. 103 as being unpatentable over Kristal (“SYSTEMS AND METHODS FOR ASSESSING OUTCOMES OF THE COMBINATION OF PREDICTIVE OR DESCRIPTIVE DATA MODELS”, U.S. Publication Number: 20220344060 A1), in view of Zadeh (“METHODS AND SYSTEMS FOR APPLICATIONS FOR Z-NUMBERS”, U.S. Publication Number: 20140201126 A1),in view of Nair (“SYSTEMS AND METHODS FOR PREDICTIVE EARLY STOPPING IN NEURAL NETWORK TRAINING”, U.S. Publication Number: 20200372342 A1),in view of Lackey (“COMPUTING ESTIMATED CLOSEST CORRELATION MATRICES”, U.S. Publication Number: 20240004953 A1).
Regarding Claim 1,
Kristal teaches,
A method for providing a synthetic neural data model, the method being implemented by at least one processor, the method comprising:
(Kristal [0188] The procedure for generating synthetic exponentially modified data
Kristal [0049] modelling methods applied in these domains span the breadth of statistical and mathematical knowledge... to ensemble classifiers and the very latest deep-learning neural networks
Kristal [0068] the computing devices 102, 104 can include a processor device
Kristal [0122] provide real-time and static testing of model applicability, which can be used to determine which data models/predictive algorithms (or models) should be included when multiple such tools are available, and which data sets (e.g., two different financial datasets that differ in the variables measured in that dataset).
Kristal [Claim 1] a processor device)
appending, by the at least one processor, at least one agent to the model, the at least one agent
(Kristal [0003] it may be advantageous to combine multiple models ....combining multiple models may not provide any advantages
Kristal [0011] an example of fusing (or concatenating) together a first data model (which is first raw data) with a second data model (which is second raw data).
Kristal [0049] optimally combine prediction and/or classification algorithms directly)
relating to a software component that sends at least one request to the model based on a predetermined timestep;
(Kristal [0247] programming or engineering techniques to produce software
Kristal [0151] allow one to determine the best model or models under current, shifting conditions, and use this to assign probabilities to specific bounds being crossed, and thus triggering stop/loss orders or buy triggers.... to set a trigger, this increases flexibility, and provides a trading edge.
Kristal [0059] which can consider more options in similar time (e.g., in financial options, you can not only consider spreads within a group, e.g., financials, but you can consider all available futures contracts (asset, time, strike, and the like) essentially simultaneously).
Kristal [0087] plurality of correlation groups each have a uniform interval of correlation, across the entire span of correlation (e.g., from −1 to 1). For example, a number of “slices” or correlation groups (or intervals) can be determined)
assigning, by the at least one processor, a fixed grid to the model, the fixed grid including a tick size that relates corresponds to a fixed granularity
(Kristal [0091] In some configurations, the curves can be generated, for each correlation slice (or interval)
Kristal [0092] this can include storing the coordinates of the curve, a function defined by the curve, an association between the specific curve and the correlation interval
Kristal [0093] depending on the desired resolution, holding one accuracy value relatively constant (e.g., a substantially small accuracy interval) can provide an allowable accuracy range to be used for the other accuracy (e.g., that the other accuracy must be within for a recommendation for fusion), and vice versa. These ranges, intervals, etc., can be appropriately stored, such as in a look-up table
Kristal [0059] which can consider more options in similar time (e.g., in financial options, you can not only consider spreads within a group, e.g., financials, but you can consider all available futures contracts (asset, time, strike, and the like) essentially simultaneously).
Kristal [0087] plurality of correlation groups each have a uniform interval of correlation, across the entire span of correlation (e.g., from −1 to 1). For example, a number of “slices” or correlation groups (or intervals) can be determined
Kristal [0199] Whenever the size of a dataset is fixed)
in which consecutive values are spaced apart by a distance that corresponds to the tick size;
(Kristal [0199] Whenever the size of a dataset is fixed
Kristal [0230] thus would have sequence elements [1−M, 2−M, . . . , 0, . . . , N−M].
Kristal [0091] In some configurations, the curves can be generated, for each correlation slice (or interval)
Kristal [0093] such as in a look-up table in a computer readable medium so that these values are concrete (e.g., fixed), easily recalled, and easily comparable.)
obtaining, by the at least one processor, historical data; inputting, by the at least one processor, the model and the historical data into a neural network; and
(Kristal [0102] In some non-limiting examples, the order of evaluation of pairs of the data models can be implemented in a predetermined order, or an order based on specific operational conditions (e.g., based on clustering, correlation, and the like), or conditions selected by a user (e.g., via a user input).
Kristal [0067] the computing device 102 can store data in and receive data from (e.g., calculated curves for storage in a database, look-up table, and the like) the server
Kristal [0092] data can be stored in a lookup table for easy recall or comparisons of the data.
Kristal [0049] modelling methods applied in these domains... to ensemble classifiers and the very latest deep-learning neural networks)
beginning, by the at least one processor via the neural network, training of a neural network extension of the model
(Kristal [0050] occur at three conceptually different levels, pre-training (typically called data- or feature-level fusion), post-training (known by many names, e.g., system or model level fusion)... datasets for training, optimization, and testing
Kristal [0049] modelling methods applied in these domains... to ensemble classifiers and the very latest deep-learning neural networks
Kristal [0085] metrics can be designed to optimize either high or low values
Kristal [0132] can switch between individual models and fused systems, always optimizing given parameters)
and using an optimizer
(Kristal [0049] datasets for training, optimization, and testing.)
Kristal does not teach generating, by the at least one processor, a model that is configured to simulate an electronic communication network; calibrating, by the at least one processor, each of the at least one agent by using a calibration data set; generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate training loss; identifying, by the at least one processor via the neural network, a multivariate Gaussian distribution mixture that characterizes the electronic communication network; assessing, by the at least one processor, a validation loss parameter of the model and the neural network extension: plotting a neural network extension learning curve for the validation loss parameter; and stopping, by the at least one processor, training of the neural network extension when the validation loss parameter begins to increase.
Zadeh teaches,
generating, by the at least one processor, a model that is configured to simulate an electronic communication network; that characterizes the electronic communication network
(Zadeh [2907] In one embodiment, the system above is used for neural network simulations or actual hardware implementation of that, with nodes and layers, using devices on substrate in different layers of semiconductor structure, to connect to other devices or terminals or metal contacts.... to represent and mimic neural networks (with nodes and layers, plus bias feed).)
calibrating, by the at least one processor, each of the at least one agent by using a calibration data set;
(Zadeh [1136] a software agent
Zadeh [1138] Database comprises a collection of agent-controlled modules and submodules, each of which contains rules drawn from various fields and various modalities of generalized constraints.
Zadeh [1121] These examples point to an important aspect of precisiation. Specifically, to precisiate p, we have to precisiate or, equivalently, calibrate its lexical constituents.
Zadeh [1403] The correlation function may be normalized for amplitude, using correlation coefficient (e.g. for changes in size or rotation).
Zadeh [1687] One advantage of reusing the predetermined normalized categories is the reduction in number of calculation...will involve additional overlap determination involving various normalized category sets
Zadeh [0796] based on the source's own knowledge base (or database) and processor)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the calibration and normalization teachings of Zadeh with “calibration metrics or normalization factor” (Zadeh [1885]). The modification would have been obvious, because it is merely applying a known technique (i.e. calibration and normalization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. as “it is sensitive to noisy data and outliers…it is less susceptible to the ‘overfitting’ problem” Zadeh [2195])
Zadeh does not teach generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate training loss; identifying, by the at least one processor via the neural network, a multivariate Gaussian distribution mixture…; assessing, by the at least one processor, a validation loss parameter of the model and the neural network extension: plotting a neural network extension learning curve for the validation loss parameter; and stopping, by the at least one processor, training of the neural network extension when the validation loss parameter begins to increase.
Nair teaches,
assessing, by the at least one processor, a validation loss parameter of the model and the neural network extension: plotting a neural network extension learning curve for the validation loss parameter; and stopping, by the at least one processor, training of the neural network extension when the validation loss parameter begins to increase;
(Nair [Abstract] may train neural networks (NNs) and determine when to stop training
Nair [0075] In operation 460, if the current or actual training loss is less than a historic minimum training loss the wait value is not increased,...an integer wait value is used to determine a period of waiting or “patience” where if no improvement is seen over the period, training is stopped.
Nair [0093] that actual loss 602 for a NN converges with predicted final loss 600 as the number of epochs increase....a stopping algorithm ...has been applied
Nair [0049] time series features may be input along with hyperparameters and graph features ...a graph representing the NN ... Extracted timeseries features, hyperparameters (e.g. batch size, learning rate, number of training samples, etc.),... or other features such as the number of traininable parameters present in the mode may be analyzed.)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the training termination teachings of Nair to “determine when to stop training” (Nair [Abstract]). The modification would have been obvious, because it is merely applying a known technique (i.e. training termination) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “to not waste computing or other resources.” Nair [Abstract])
Nair does not teach generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate training loss; identifying, by the at least one processor via the neural network, a multivariate Gaussian distribution mixture…;
Lackey teaches,
generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate training loss;
(Lackey [0001] A correlation matrix is a matrix in which the elements are correlation coefficients between pairs of random variables. Correlation matrices are symmetric and positive definite, and each element along the main diagonal of a correlation matrix is equal to 1.
Lackey [0042] may be further configured to compute a value of a loss function 64 for the machine learning model
Lackey [Abstract] The one or more processors may be further configured to train a machine learning model using the training data set.)
identifying, by the at least one processor via the neural network, a multivariate Gaussian distribution mixture…;
(Lackey [0014] input correlation coefficients (e.g., user-provided coefficients, empirically generated coefficients, or coefficients derived from a copula)
Lackey [0027] input correlation coefficients ρij based at least in part on a copula 24. The copula 24 may be a copula 24 whose dimension equals the number of the plurality of marginal distributions 20. For example, the copula 24 may be a Gaussian copula.
Lackey [0041] the machine learning model 60 may be a deep neural network)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the correlation matrices of Lackey with “A correlation matrix is a matrix in which the elements are correlation coefficients between pairs of random variables” (Lackey [0001]). The modification would have been obvious, because it is merely applying a known technique (i.e. correlation matrices) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “indicate correlations between a plurality of pairs 22 of correlated random variables that have underlying marginal distributions.” Lackey [0020])
Regarding Claim 5,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal teaches,
wherein training the neural network extension further comprises: training, by the at least one processor, the neural network extension by using the optimizer;
(Kristal [0050] occur at three conceptually different levels, pre-training (typically called data- or feature-level fusion), post-training (known by many names, e.g., system or model level fusion)... datasets for training, optimization, and testing
Kristal [0049] modelling methods applied in these domains... to ensemble classifiers and the very latest deep-learning neural networks
Kristal [0085] metrics can be designed to optimize either high or low values
Kristal [0132] can switch between individual models and fused systems, always optimizing given parameters)
Kristal does not teach plotting, by the at least one processor, a learning curve for the neural network extension.
Zadeh teaches,
plotting, by the at least one processor, a learning curve for the neural network extension.
(Zadeh [2223] It also can be used in or with any learning theory, e.g. VC theory (including VC dimension and VC bound), Bias-Variance theory (for learning curve analysis)
Zadeh [2488] a set of presentation tools, such as graphs or 2-D Cartesian drawings (for Y versus X axis) or tables, to present the raw facts in a presentable format or modified format, as required
Zadeh [0553] can be used to draw curves and plot functions.
Zadeh [0564] Note. Originally, the term “extension principle” was employed to describe a rule which serves to extend the domain of definition of a function
Zadeh [1381] by machine or processor, automatically, with a training module, such as a neural network)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the calibration and normalization teachings of Zadeh with “calibration metrics or normalization factor” (Zadeh [1885]). The modification would have been obvious, because it is merely applying a known technique (i.e. calibration and normalization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. as “it is sensitive to noisy data and outliers…it is less susceptible to the ‘overfitting’ problem” Zadeh [2195])
Regarding Claim 6,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 5 as described earlier.
Kristal does not teach initializing, by the at least one processor, at least one neural network weight by using an initialization function to ensure uniform distribution, wherein the at least one neural network weight relates to a parameter within the neural network extension that transforms input data within a hidden layer of the neural network extension.
Zadeh teaches,
initializing, by the at least one processor, at least one neural network weight by using an initialization function to ensure uniform distribution,
(Zadeh [1328] machines can initialize the new search links and ... relationships, automatically, without any human intervention or input.
Zadeh [1722] A Boltzmann machine refers to a type of stochastic recurrent neural network, where the probability of the state is based on an energy function defined based on the weights/biases associated with the units and the state of such units
Zadeh [2448] use restricted-centered theory of reasoning and computation in an environment of uncertainty and imprecision (also called RRC)...restriction can be probabilistic (e.g. "X has a uniform probability distribution."))
wherein the at least one neural network weight relates to a parameter within the neural network extension that transforms input data within a hidden layer of the neural network extension.
(Zadeh [1722] A Boltzmann machine refers to a type of stochastic recurrent neural network, where the probability of the state is based on an energy function defined based on the weights/biases associated with the units and the state of such units
Zadeh [1745] In one embodiment, elastic distortions (as well as affine transformations) are used to expand the size and variety of the training set, e.g., when the training set is produced from a model
Zadeh [1743] In one embodiment, the training is done one hidden layer at the time (e.g., until H.sup.(3)). In one embodiment, the training of hidden layers is done unsupervised)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the calibration and normalization teachings of Zadeh with “calibration metrics or normalization factor” (Zadeh [1885]). The modification would have been obvious, because it is merely applying a known technique (i.e. calibration and normalization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. as “it is sensitive to noisy data and outliers…it is less susceptible to the ‘overfitting’ problem” Zadeh [2195])
Regarding Claim 9,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal teaches,
wherein the model includes at least one from among a machine learning model, a mathematical model, a process model, and a data model.
(Kristal [0076] models in which the inner workings are either not obvious to the creators of the algorithms (e.g., machine learning, AI, deep learning classifiers) or are deliberately loaded into a system
Kristal [0169] choose between leveraging a look-up table or mathematical calculations
Kristal [0179] A breakthrough mathematical framework that accurately forecasts the utility of combining predictive or descriptive mathematical models
Kristal [Absract] for fusing the first data model with the second data model.)
Claim 10 is rejected on the same basis as Claim 1.
Claim 14 is rejected on the same basis as Claim 5.
Claim 15 is rejected on the same basis as Claim 6.
Claim 18 is rejected on the same basis as Claim 9.
Claim 19 is rejected on the same basis as Claim 1.
Regarding Claim 21,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal teaches,
reducing a dimensionality of the model by determining a default difference between values of consecutive non-empty variable levels,
(Kristal [0073] due to the loss of distributional information and the reduced effect of valid outliers—outliers in the correct direction/direction desired
Kristal [0103] can determine a difference between the first and second accuracies for the given data model pair.
Kristal [0107] data model pairs, currently, and as a default future
Kristal [0199] Whenever the size of a dataset is fixed
Kristal [0230] thus would have sequence elements [1−M, 2−M, . . . , 0, . . . , N−M]. )
wherein the default difference is the distance of the tick size.
(Kristal [0091] In some configurations, the curves can be generated, for each correlation slice (or interval)
Kristal [0093] depending on the desired resolution, holding one accuracy value relatively constant (e.g., a substantially small accuracy interval) can provide an allowable accuracy range to be used for the other accuracy ... These ranges, intervals, etc., can be appropriately stored, such as in a look-up table)
Claim 22 is rejected on the same basis as Claim 21.
Regarding Claim 23,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal teaches,
utilizing the …correlation matrix
(Kristal [0055] diversity between these systems using common correlation metrics (e.g., Pearson correlation) were measured
Kristal [0206] a pairwise within-class correlation measurement and fusions
Kristal [0067] calculated curves for storage in a database, look-up table, and the like
Kristal [0092] This data can be stored in a lookup table for easy recall or comparisons of the data)
Kristal does not teach symmetrical correlation matrix; to obtain errors for each training example; and generating the validation loss parameter from a sum of the errors.
Nair teaches,
to obtain errors for each training example;
(Nair [Abstract] a model trained using training data from other NNs may return a probability of improvement in the loss of the NN
Nair [0004] and an optimization procedure that minimizes the training loss
Nair [0008] generates an error e.g., a loss
Nair [0018] may reduce model training time by an average of 20%, with an average error rate of 4%)
and generating the validation loss parameter from a sum of the errors.
(Nair [0075] In operation 460, if the current or actual training loss is less than a historic minimum training loss the wait value is not increased,...an integer wait value is used to determine a period of waiting or “patience”
Nair [0093] that actual loss 602 for a NN converges with predicted final loss 600 as the number of epochs increase)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the errors for loss parameter teachings of Nair to “generates an error e.g., a loss” (Nair [0008]). The modification would have been obvious, because it is merely applying a known technique (i.e. errors for loss parameter) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “actual loss 602 for a NN converges with predicted final loss.” Nair [0093])
Nair does not teach symmetrical correlation matrix
Lackey teaches,
symmetrical correlation matrix
(Lackey [0001] A correlation matrix is a matrix in which the elements are correlation coefficients between pairs of random variables. Correlation matrices are symmetric and positive definite)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the correlation matrices of Lackey with “A correlation matrix is a matrix in which the elements are correlation coefficients between pairs of random variables” (Lackey [0001]). The modification would have been obvious, because it is merely applying a known technique (i.e. correlation matrices) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “indicate correlations between a plurality of pairs 22 of correlated random variables that have underlying marginal distributions.” Lackey [0020])
Claim 24 is rejected on the same basis as Claim 23.
Claims 2, 11, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Kristal, Zadeh, Nair, and Lackey in view of Jang (“GENERATION OF TIME-INTERVAL-SPECIFIC SUPPORT VECTOR MACHINE”, U.S. Patent Number: US 12293412 B1).
Regarding Claim 2,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal does not teach wherein the at least one model includes support for a plurality of requests, each of the plurality of requests is tracked by using the at least one model with a first-in-first-out mechanism, and each of the plurality of requests includes a corresponding requestor.
Jang teaches,
wherein the at least one model includes support for a plurality of requests, each of the plurality of requests is tracked by using the at least one model with a first- in-first-out mechanism, and each of the plurality of requests includes a corresponding requestor
(Jang [Col 37, Lines 6-10] a first-in/first-out (FIFO) matching algorithm, also referred to as a “Price Time” algorithm, considers each identified order sequentially in accordance with when the identified order was received.)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the first-in-first-out mechanism of Jang that “allocates any still remaining quantity of the incoming order using the FIFO or pro-rata algorithms” (Jang [Col 37, Lines 37-39]). The modification would have been obvious, because it is merely applying a known technique (i.e. first-in-first-out mechanism) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “FIFO generally rewards the first trader to place an order at a particular price and maintains this reward indefinitely.” Jang [Col 40, Lines 30-33])
Claim 11 is rejected on the same basis as Claim 2.
Claim 20 is rejected on the same basis as Claim 2.
Claims 4, 8, 13, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Kristal, Zadeh, Nair, and Lackey in view of Brookfield (“SYSTEM AND METHOD FOR IMPLEMENTING A DYNAMIC SIMULATION SYSTEM”, U.S. Publication Number: 20160225085 A1).
Regarding Claim 4,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal does not teach wherein the calibration data set relates to level two limit order book data, the calibration data set including at least one limit order book snapshot for each of a plurality of predetermined times.
Zadeh teaches,
wherein the calibration data set; the calibration data set
(Zadeh [1121] These examples point to an important aspect of precisiation. Specifically, to precisiate p, we have to precisiate or, equivalently, calibrate its lexical constituents.
Zadeh [1403] The correlation function may be normalized for amplitude, using correlation coefficient (e.g. for changes in size or rotation).
Zadeh [1687] One advantage of reusing the predetermined normalized categories is the reduction in number of calculation...will involve additional overlap determination involving various normalized category sets)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the calibration and normalization teachings of Zadeh with “calibration metrics or normalization factor” (Zadeh [1885]). The modification would have been obvious, because it is merely applying a known technique (i.e. calibration and normalization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. as “it is sensitive to noisy data and outliers…it is less susceptible to the ‘overfitting’ problem” Zadeh [2195])
Zedah does not teach relates to level two limit order book data; including at least one limit order book snapshot for each of a plurality of predetermined times.
Brookfield teaches,
relates to level two limit order book data;
(Brookfield [0003] bid quantities and/or ask quantities of a tradable object to enable the user to determine a market depth
Brookfield [0069] Market data may be organized according to a plurality of levels (or price tiers), the depth and/or detail of market data. For example, each successive level (e.g., level one, level two, level three, etc.)
Brookfield [0112] a type of order 744 (e.g., a limit order, etc.)
Brookfield [0055] includes an order book)
including at least one limit order book snapshot for each of a plurality of predetermined times.
(Brookfield [0077] Live and/or saved market data may be provided as a data stream, snapshot
Brookfield [0112] a type of order 744 (e.g., a limit order, etc.)
Brookfield [0055] includes an order book
Brookfield [0083] market data has been recorded or otherwise saved for a specified duration from a specified or predetermined start time to a specified or predetermined end time.)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the electronic trading system of Brookfield that “receives information about a market, such as prices and quantities, from the electronic exchange… attempts to match quantity of an order with quantity of one or more contra-side order” (Brookfield [0002]). The modification would have been obvious, because it is merely applying a known technique (i.e. electronic trading system) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “provide a trading interface to enable a user to monitor the information about the market and execute trades via the electronic exchange.” Brookfield [0003])
Regarding Claim 8,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal does not teach predicting, by the at least one processor via the model, at least one change in volume at each of a plurality of levels; and splitting, by the at least one processor, an output into at least one component output for transmitting to the electronic communication network.
Zadeh teaches,
splitting, by the at least one processor, an output into at least one component output for transmitting to the electronic communication network.
(Zadeh [2946] information is uploaded/pulled/pushed separately to the server. In one embodiment, the server transmits
Zadeh [1641] communication between different units, devices, or modules are done by wire, cable, fiber optics, wirelessly, WiFi, Bluetooth, through network, Internet)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the calibration and normalization teachings of Zadeh with “calibration metrics or normalization factor” (Zadeh [1885]). The modification would have been obvious, because it is merely applying a known technique (i.e. calibration and normalization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. as “it is sensitive to noisy data and outliers…it is less susceptible to the ‘overfitting’ problem” Zadeh [2195])
Zadeh does not teach predicting, by the at least one processor via the model, at least one change in volume at each of a plurality of levels;
Brookfield teaches,
predicting, by the at least one processor via the model, at least one change in volume at each of a plurality of levels;
(Brookfield [0072] Other approaches model an entire market depth and user orders within that depth, with an estimation of an order's place in a queue at the order's price tier in the market.
Brookfield [0101] maintains a liquidity impairment delta at each price tier for a duration of the replay, which can reduce an effective volume at that tier to zero at minimum.
Brookfield [0026] Market depth refers to quantities available in the market. For example, the market depth includes the quantity at various price levels )
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the electronic trading system of Brookfield that “receives information about a market, such as prices and quantities, from the electronic exchange… attempts to match quantity of an order with quantity of one or more contra-side order” (Brookfield [0002]). The modification would have been obvious, because it is merely applying a known technique (i.e. electronic trading system) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “provide a trading interface to enable a user to monitor the information about the market and execute trades via the electronic exchange.” Brookfield [0003])
Claim 13 is rejected on the same basis as Claim 4.
Claim 17 is rejected on the same basis as Claim 8.
Claims 25 and 26 are rejected under 35 U.S.C. 103 as being unpatentable over Kristal, Zadeh, Nair, and Lackey in view of Wu (“COMMUNICATION LOAD BALANCING VIA META MULTI-OBJECTIVE REINFORCEMENT LEARNING”, U.S. Publication Number: 20230084465 A1).
Regarding Claim 25,
Kristal, Zadeh, Nair, and Lackey teach the synthetic neural data model of Claim 1 as described earlier.
Kristal does not teach utilizing the at least one request to stabilize the electronic communication network of the model.
Wu teaches,
utilizing the at least one request to stabilize the electronic communication network of the model.
(Wu [0102] then applies network state 2-16 as an input to πnew and obtains parameters 1-9 for balancing traffic flowing through
Wu [0113] The multi-objective approach of embodiments is more effective to improve cellular network performance than single policy and traditional rule-based approaches.)
It is prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the synthetic neural data model of Kristal to incorporate the network stabilization of Wu to “balancing traffic flowing through” (Wu [0102]). The modification would have been obvious, because it is merely applying a known technique (i.e. network stabilization) to a known concept (i.e. synthetic neural data model) ready for improvement to yield predictable result (i.e. “improve cellular network performance” (Wu [0113])
Claim 26 is rejected on the same basis as Claim 25.
Response to Remarks
Applicant's arguments filed on March 25, 2026, have been fully considered and Examiner’s remarks to Applicant’s amendments follow.
Response Remarks on Claim Rejections - 35 USC § 101
The Applicant states:
“Applicant respectfully submits that these newly added features not only integrate the alleged abstract idea into a practical application, but that they also amount to significantly more than the alleged abstract idea. Additionally, Applicant respectfully submits that the claims can no longer reasonably be considered to be directed to a fundamental economic principle or practice and/or commercial or legal interaction because the claims no longer recite any element that can reasonably be considered to be a commercial interaction, a legal interaction, or a fundamental economic principle or practice."
Examiner responds:
The limitations, under their broadest reasonable interpretation, continue to cover performance of certain methods of organizing human activity. Specific instances include instructions for “obtaining…historical (order book) data” recites a fundamental economic principles or practice and/or commercial or legal interactions. Furthermore, the Specification reads:
[0072] the electronic communication network may relate to a computerized system that automatically matches buy and sell orders for securities in a financial market.
[0007] the at least one model may include support for a plurality of orders that relate to at least one from among a market order, a limit order, and a cancel order, the support may enable the plurality of orders on a bid and ask side.
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a fundamental economic, commercial, or financial action, principle, or practice then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas.
Furthermore, the limitations, under their broadest reasonable interpretation, cover performance as Mathematical Concepts. Specific instances include instructions that "providing a synthetic neural data model" and "a model that is configured to simulate" and "assigning...a fixed grid to the model" and "calibrating...using a calibration data set" and "using an optimizer to generate a symmetrical correlation matrix that comprises a diagonal of ones and to utilize the symmetrical correlation matrix to evaluate … loss " and “a multivariate Gaussian distribution mixture” recite Mathematical Concepts. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as mathematical relationships or mathematical calculations then it falls within the “Mathematical Concepts” grouping of abstract ideas.
The Applicant states:
“Applicant respectfully submits that the newly added independent claim features nevertheless integrate such idea into a practical application by applying the claimed concept(s) to the programming and training of artificial intelligence and machine learning models (namely, synthetic neural data models), because the claimed application of ideas improves artificial intelligence and machine learning technology, for example, by providing such technology with a process that provides optimum artificial intelligence and machine learning model training…In other words, the claimed application of ideas improves artificial intelligence and machine learning models through retraining until such retraining no longer improves the model, which thereby provides optimum artificial intelligence and machine learning model training."
Examiner responds:
Examiner contends the Applicant utilizes conventional and off-the-shelf techniques and components. The Specification reads:
[0090] In another exemplary embodiment, consistent with present disclosures, the neural network extension may be trained by using an optimizer such as, for example, an ADAM optimizer with learning rate 10-3 up to epoch 110, then 10-4, and a minibatch size of 100. The network weights may be initialized by using an initialization function such as, for example, a XAVIER initialization that is associated with the uniform distribution.
An ADAM optimizer is commonly used for retraining optimization. Devapatla ("Getting to Know Adam Optimization: A Comprehensive Guide", Tech Mahindra, March 30, 2023) writes, "One of the most popular optimization techniques used in training neural networks is the Adam optimizer.…Adam optimization is a gradient descent-based optimization algorithm introduced by Diederik P. Kingma and Jimmy Ba in 2014. Adam stands for Adaptive Moment Estimation, which describes the optimizer's method to update weights during training. The basic idea behind Adam optimization is to adjust the learning rate adaptively for each parameter in the model based on the history of gradients calculated for that parameter. This helps the optimizer converge faster and more accurately than fixed learning rate methods like stochastic gradient descent (SGD)" and Rath ("Adam Optimizer for Deep Learning Optimization", June 29, 2020) teaches, "Stochastic Gradient Descent (I will refer to it as SGD from here on) has played a major role in many successful deep learning projects and research experiments. But SGD has its own limitations as well. The requirement of excessive tuning of the hyperparameters is one of them. Recently the Adam optimization algorithm has gained a lot of popularity. Adam was developed by Diederik P. Kingma, Jimmy Ba in 2014 and works well in place of SGD."
The focus of the claims is not on an improvement in machine learning as a tool but on certain independently abstract ideas that use machine learning as a tool.
Nothing in the claims, understood in light of the specification, requires anything other than “merely applying” off-the-shelf, conventional computer, machine learning, and optimization technology for gathering, synthesizing, sending, and presenting the desired information. See MPEP 2106.05(d) well-understood, routine, and conventional.
In the absence of unexpected results, changes or alteration of sequence do not make for a patentable invention, see Ex parte Rubin, 128 USPQ 440 (Bd. App. 1959) ; In re Burhans, 154 F.2d 690, 69 USPQ 330 (CCPA 1946); In re Gibson, 39 F.2d 975, 5 USPQ 230 (CCPA 1930)
The Applicant states:
“Applicant respectfully submits that these features cannot reasonably be considered to be abstract because they are directed to an inventive concept for improving neural network training… See the ARP's decision on request for rehearing in Ex parte Desjardins et al."
Examiner responds:
In Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision) the specification identified the improvement to machine learning technology by explaining how the machine learning model is trained to learn new tasks while protecting knowledge about previous tasks to overcome the problem of “catastrophic forgetting,” and that the claims reflected the improvement identified in the specification. Indeed, enumerated improvements identified in the Desjardins specification included disclosures of the effective learning of new tasks in succession in connection with specifically protecting knowledge concerning previously accomplished tasks; allowing the system to reduce use of storage capacity; and the enablement of reduced complexity in the system. Such improvements were tantamount to how the machine learning model itself would function in operation and therefore not subsumed in the identified mathematical calculation.
Applicant’s invention incorporates no similar details and is not analogous to Ex Parte Desjardins. Applicant’s use of generic components function as designed with no unexpected results. Again, the Applicant’s Specification reads:
[0090] In another exemplary embodiment, consistent with present disclosures, the neural network extension may be trained by using an optimizer such as, for example, an ADAM optimizer with learning rate 10-3 up to epoch 110, then 10-4, and a minibatch size of 100. The network weights may be initialized by using an initialization function such as, for example, a XAVIER initialization that is associated with the uniform distribution.
An ADAM optimizer is commonly used for neural network retraining optimization.
Therefore, the rejection under 35 USC § 101 remains.
Response Remarks on Claim Rejections - 35 USC § 103
Applicant's amendments required the application of new/additional prior art.
New prior art includes:
Lackey (“COMPUTING ESTIMATED CLOSEST CORRELATION MATRICES”, U.S. Publication Number: 20240004953 A1).
Applicant’s remarks regarding the rejection made under 35 USC § 103 are rendered moot by the introduction of additional prior art.
Therefore, the rejection under 35 USC § 103 remains.
Prior Art Cited But Not Applied
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Devapatla ("Getting to Know Adam Optimization: A Comprehensive Guide", Tech Mahindra, March 30, 2023) writes, "One of the most popular optimization techniques used in training neural networks is the Adam optimizer.…Adam optimization is a gradient descent-based optimization algorithm introduced by Diederik P. Kingma and Jimmy Ba in 2014. Adam stands for Adaptive Moment Estimation, which describes the optimizer's method to update weights during training. The basic idea behind Adam optimization is to adjust the learning rate adaptively for each parameter in the model based on the history of gradients calculated for that parameter. This helps the optimizer converge faster and more accurately than fixed learning rate methods like stochastic gradient descent (SGD)"
Rath ("Adam Optimizer for Deep Learning Optimization", June 29, 2020) teaches, "Stochastic Gradient Descent (I will refer to it as SGD from here on) has played a major role in many successful deep learning projects and research experiments. But SGD has its own limitations as well. The requirement of excessive tuning of the hyperparameters is one of them. Recently the Adam optimization algorithm has gained a lot of popularity. Adam was developed by Diederik P. Kingma, Jimmy Ba in 2014 and works well in place of SGD."
Lange (“SYSTEMS AND METHODS FOR CROWDSOURCING OF ALGORITHMIC FORECASTING”, U.S. Publication Number: 20150206246 A1) proposes new computational technologies generating systematic investment portfolios by coordinating forecasting algorithms contributed by researchers are provided. … the algorithm selection system performs a batch of tests that selects the best developed algorithms, updates the list of open challenges and translates those scientific forecasts into financial predictions. The algorithm controls for the probability of backtest overfitting and selection bias, thus providing for a practical solution to a major flaw in computational research involving multiple testing. Third, the incubation system verifies the reliability of those selected algorithms. Fourth, the portfolio management system uses the selected algorithms to execute investment recommendations. A dynamically optimal portfolio trajectory is determined by a quantum computing solution to combinatorial optimization representation of the capital allocation problem.
Balakrishnan (“FEDERATED LEARNING OPTIMIZATIONS”, U.S. Publication Number: 20230177349 A1) provides an initial set of weights for a global machine learning (ML) model to be transmitted a set of client compute nodes of the edge computing network; process Hessians computed by each of the client compute nodes based on a dataset stored on the client compute node; evaluate a gradient expression for the ML model based on a second dataset and an updated set of weights received from the client compute nodes; and generate a meta-updated set of weights for the global model based on the initial set of weights, the Hessians received, and the evaluated gradient expression.
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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHINEDU EKECHUKWU whose telephone number is (571) 272-4493. The examiner can normally be reached on Mon-Fri 9 AM ET to 3:30 PM ET.
Examiner interviews are available via telephone 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, Christine Tran, can be reached on (571) 272-8103. 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.
/C.E./Examiner, Art Unit 3695
/CHRISTINE M Tran/Supervisory Patent Examiner, Art Unit 3695