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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 10/31/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
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.
Claim(s) 1 is/are rejected under 35 U.S.C. 103 as being unpatentable over Akhtar (US 11132853 B1) in view of Ambikavathi (IJITEE 2020, “Tuning Random Forest Parameters using Simulated Annealing for Intrusion Detection”).
Regarding claim 1, Akhtar teaches a method for remotely diagnosing faults in a powertrain of an electric vehicle (col 1 last para, col 19 1st para wherein “The management server can receive the data from devices for many vehicles, many fleets, and over extended periods of time. The management server can aggregate and analyze the received data”; “The vehicle battery data can be for an electric vehicle, a hybrid vehicle (such as a plug-in hybrid electric vehicle)”, the method comprising:
collecting real-time data from the powertrain of the electric vehicle; transmitting the real-time data to a cloud platform (col 1 last para, col 4 last para wherein “The management server can receive the data from devices for many vehicles, many fleets, and over extended periods of time. The management server can aggregate and analyze the received data”; “A management server can provide an analysis graphical user interface that allows a user to review the vehicle metrics, benchmarks, and/or summary data in substantially real-time”).
ranking the real-time data in terms of a feature importance using a random forest model (col 42 2nd para wherein “For example, the management server 140 can use a random forest model to determine feature importance, which can be used for ranking the attributes”);
deploying the random forest model on the cloud platform for real-time fault classification and diagnosis (col 29 2nd para wherein the fault is classified as types and diagnosed by the server using the random forest model; “Moving to block 721, an event type associated with the detected safety-related event may be determined. In particular, the management server 140 may first determine an event type associated with the detected safety-related event. The event type may then be used to select one or more event models to be tested or updated based on the event data”); and
using the cloud platform to provide real-time fault alerts and diagnostic reports for the powertrain of the electric vehicle (col 4 last para, col 19 1st-3rd para wherein “A management server can provide an analysis graphical user interface that allows a user to review the vehicle metrics, benchmarks, and/or summary data in substantially real-time”).
However, Akhtar fails to teach optimizing hyperparameters of the random forest model using a simulated annealing algorithm and using the optimized random forest model for fault classification and diagnosis.
Akhtar further teaches hyperparameters of random forest model (col 42 3rd para wherein “Random forest models consist of a number of decision trees. Every node in the decision trees can be a condition on a single feature, designed to split the dataset into two so that similar response values end up in the same set”).
Ambikavathi teaches optimizing hyperparameters of the random forest model using a simulated annealing algorithm and using the optimized random forest model for fault diagnosis (Abstract, page 355, section B wherein “Simulated Annealing is a randomized local search algorithm. It is used to find optimal parameters in a discrete search space with large number of combination of hyper parameters”).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Akhtar’s teachings of using random forest model to classify and diagnose faults to incorporate Ambikavathi’s teachings of optimizing hyperparameters of the random forest model using a simulated annealing algorithm and using the optimized random forest model for fault diagnosis in order to optimize hyperparameters of the random forest model using a simulated annealing algorithm and using the optimized random forest model for fault classification and diagnosis. Doing so would improve the fault detection using hyperparameters of random forest model (Ambikavathi, abstract, “Therefore Simulated Annealing (SA) is utilized for tuning these hyper parameters of RF which leads to improve detection accuracy and efficiency of IDS”).
Claim(s) 2 is/are rejected under 35 U.S.C. 103 as being unpatentable over Akhtar (US 11132853 B1), Ambikavathi (IJITEE 2020, “Tuning Random Forest Parameters using Simulated Annealing for Intrusion Detection”) in view of Alabdullatif (US 20220188665 A1).
Regarding claim 2, Akhtar teaches wherein the random forest model performs the real-time fault classification and diagnosis based on a training set (col 3 5th para wherein “In various embodiments, training the random forest model can further include: providing the plurality of segmentation attributes as feature input and a plurality of metrics as label input to the random forest model”), an importance Ij of a feature j in the training set is calculated (col 42 2nd-3rd para wherein “Building the model, such as a tree-based model, is a way to evaluate how important each feature/attribute in relation to the metric that is trying to be predicted. For example, the management server 140 can use a random forest model to determine feature importance, which can be used for ranking the attributes. Example methods for determining feature importance can include (i) mean decrease impurity and (ii) mean decrease accuracy”; “Thus, when the management server trains the tree model, the management server 140 can compute how much each feature decreases the weighted impurity in the tree”).
However, Akhtar fails to teach the importance of the feature in the training set is calculated by Formula I: Ij=1/T∑t=1TΔGini(t,j) Formula I wherein, ΔGini(t, j) being a reduced value of Gini index caused by feature j in decision tree t, and T being a total number of decision trees.
Alabdullatif teaches the importance of the feature in the training set is calculated by Formula I: Ij=1/T∑t=1TΔGini(t,j) Formula I wherein, ΔGini(t, j) being a reduced value of Gini index caused by feature j in decision tree t, and T being a total number of decision trees (0037 wherein “For example, a random forest representation can characterize the feature space as a tree of nodes, each node corresponding to a candidate process variable from FIG. 2. Gini Importance or Mean Decrease in Impurity (MDI) may generally calculate each feature importance as the sum over the number of splits (across all tress) that include the feature, proportionally to the number of samples it splits. In some implementations, the Gini importance or the MDI can be calculated as the total decrease in node impurity (weighted by the probability of reaching that node (which is approximated by the proportion of samples reaching that node)) averaged over all trees of the ensemble”);
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Akhtar’s teachings of calculating an importance Ij of a feature j in the training set to incorporate Alabdullatif’s teachings of calculating the feature importance using eh above discussed relationship. Doing so would constitute combining prior art elements according to known methods of using feature importance using the provided relationship to yield predictable results of improving performance of the random forest model.
Claim(s) 3 is/are rejected under 35 U.S.C. 103 as being unpatentable over Akhtar (US 11132853 B1), Ambikavathi (IJITEE 2020, “Tuning Random Forest Parameters using Simulated Annealing for Intrusion Detection”) and Alabdullatif (US 20220188665 A1) in view of Nair (Researchgate 2023, “Simulated Annealing”).
Regarding claim 3, modified Akhtar teaches all the limitations of claim 2. Modified Akhtar also teaches wherein the simulated annealing algorithm optimizes the hyperparameters of the random forest model (Ambikavathi, Abstract, page 355, section B).
However, Akhtar fails to teach the simulated annealing algorithm performs optimization through a cooling mechanism that uses a logarithmic cooling schedule.
Nair teaches simulated annealing algorithm performs optimization through a cooling mechanism that uses a logarithmic cooling schedule (page 4, 2nd para wherein “Thus, throughout the evolution of the SA, several directions have been pursued to improve the algorithm and its usability. … One such method is simulated quenching, where the cooling schedule is a logarithmic function to allow faster cooling”).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have further modified Akhtar’s teachings that simulated annealing algorithm optimizes the hyperparameters of the random forest model to incorporate Nair’s teachings that simulated annealing algorithm performs optimization through a cooling mechanism that uses a logarithmic cooling schedule. Doing so would constitute combining prior art elements according to known methods to yield predictable results of improving performance of the simulated annealing of the hyperparameters.
Claim(s) 8-10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Akhtar (US 11132853 B1), Ambikavathi (IJITEE 2020, “Tuning Random Forest Parameters using Simulated Annealing for Intrusion Detection”) in view of Kerbs (Medium 2023, “Real-Time Anomaly Detection with AWS SageMaker & Kinesis Streams”).
Regarding claim 8, modified Akhtar teaches all the limitations of claim 1. However, Akhtar fails to teach wherein the cloud platform is Amazon Web Services (AWS) cloud platform.
Kerbs teaches the cloud platform is Amazon Web Services (AWS) cloud platform (page 2 wherein “The focus of this blog is on configuring the AWS environment for real-time machine learning inference on streaming data”).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have further modified Akhtar’s teachings of using cloud platform for processing data to incorporate Kerbs’ teachings that the cloud platform is Amazon Web Services (AWS) cloud platform. Doing so would constitute combining prior art elements according to known methods to yield predictable results of processing data using AWS.
Regarding claim 9, modified Akhtar teaches further teaches wherein the AWS platform comprises AWS SageMaker Endpoint module (Kerbs, page 2 wherein “In this blog, you will learn step-by-step on how to deploy a real-time machine learning model on AWS using SageMaker, Kinesis Streams, and Lambda functions”).
Regarding claim 10, modified Akhtar teaches further teaches wherein the AWS platform comprises AWS CloudWatch module (Kerbs, page 23 wherein “You can confirm that your real-time machine learning pipeline is functioning correctly by viewing the live CloudWatch logs for your Lambda function”).
Allowable Subject Matter
Claims 4-7 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SAGAR KC whose telephone number is (571)272-7337. The examiner can normally be reached M-F 8:30 am - 5 pm.
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, Adam Mott can be reached at (571) 270-5376. 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.
/SAGAR KC/Examiner, Art Unit 3657
/ADAM R MOTT/Supervisory Patent Examiner, Art Unit 3657