System for threshold detection using learning reinforcement

Patent metadata

Field	Value
Publication number	US-11379748-B2
Application number	US-202016901403-A
Country	US
Kind code	B2
Filing date	Jun 15, 2020
Priority date	Jun 15, 2020
Publication date	Jul 5, 2022
Grant date	Jul 5, 2022

Abstract

Systems, computer program products, and methods are described herein for dynamically determining performance benchmarking parameters based on reinforcement learning. The present invention is configured to implement the first distributed impact simulation model on an application; initiate a reinforcement learning algorithm on the application, wherein initiating further comprises receiving a performance assessment output for the one or more application parameters; initiate an optimization policy generation engine on the performance assessment output associated with the application parameters to generate an optimization to encode the performance assessment output into rewards and costs; initiate an implementation of the optimization policy on the application to maximize an aggregated reward calculated from the second portion of the first set of actions; automatically generate a second distributed impact simulation model using the second set of actions to be implemented on the application parameters; and implement the second distributed impact simulation model on the application.

First claim

What is claimed is:

1. A system for dynamic parametric modeling using learning reinforcement, the system comprising:
at least one non-transitory storage device; and
at least one processing device coupled to the at least one non-transitory storage device, wherein the at least one processing device is configured to:
electronically retrieve a first distributed impact simulation model, wherein the first distributed impact simulation model is generated using a first set of actions;
implement the first distributed impact simulation model on an application within a distributed environment, wherein implementing further comprises initiating the first set of actions on one or more application parameters;
initiate a reinforcement learning algorithm on the application, wherein initiating further comprises receiving a performance assessment output for the one or more application parameters based on at least initiating the first set of actions on the one or more application parameters;
initiate an optimization policy generation engine on the performance assessment output associated with the one or more application parameters to generate an optimization policy, wherein the optimization policy generation engine is configured to encode the performance assessment output into rewards and costs, wherein encoding further comprises assigning a cost to a first portion of the first set of actions and assigning a reward to a second portion of the first set of actions;
initiate an implementation of the optimization policy on the application, wherein initiating further comprises generating a second set of actions based on at least the optimization policy to maximize an aggregated reward calculated from the second portion of the first set of actions;
automatically generate a second distributed impact simulation model using the second set of actions to be implemented on the one or more application parameters; and
implement the second distributed impact simulation model on the application, wherein implementing further comprises initiating the second set of actions on the one or more application parameters.

2. The system of claim 1 , wherein the at least one processing device is further configured to implement the first distributed impact simulation model, wherein implementing further comprises:
electronically retrieving, from a distributed data repository, one or more data records; and
initiating an execution of the first set of actions on the one or more application parameters, wherein initiating further comprises initiating an execution of the application using the one or more data records.

3. The system of claim 1 , wherein the at least one processing device is further configured to:
electronically receive information associated with the application, wherein the information further comprises a first application type;
scan the distributed environment for one or more additional applications associated with the first application type; and
implement the second distributed impact simulation model on the one or more additional applications based on at least determining that the one or more additional applications are associated with the first application type.

4. The system of claim 1 , wherein the at least one processing device is further configured to:
electronically receive, from a computing device of a user, the first set of actions to be implemented on the one or more application parameters.

5. The system of claim 1 , wherein the at least one processing device is further configured to:
electronically receive the performance assessment output for the one or more application parameters; and
initiate a probabilistic fuzzy logic engine on the performance assessment output;
convert, using the probabilistic fuzzy logic engine, the performance assessment output to one or more output values; and
encode, using the optimization policy generation engine, the one or more output values into the rewards and costs.

6. The system of claim 1 , wherein the at least one processing device is further configured to:
initiate the first set of actions on the one or more application parameters, wherein the one or more application parameters are associated with one or more application stress test scenarios, wherein the one or more stress test scenarios comprises at least a performance stress test, an transactional stress test, a systemic stress test, and/or an exploratory stress test.

7. The system of claim 6 , wherein the at least one processing device is further configured to:
electronically receive an indication that the first distributed simulation model is associated with at least one of the one or more application stress test scenarios;
determine a first portion of the one or more application parameters associated with the at least one of the one or more application stress test scenarios; and
implement the first distributed impact simulation model on the application, wherein implementing further comprises initiating the first set of actions on the first portion of the one or more application parameters.

8. The system of claim 1 , wherein the at least one processing device is further configured to:
determine, based on at least the optimization policy, a subset of the one or more application parameters; and
implement the second distributed impact simulation model on the application, wherein implementing further comprises initiating the second set of actions on the subset of the one or more application parameters.

9. The system of claim 1 , wherein the at least one processing device is further configured to:
implement the first distributed impact simulation model, wherein implementing further comprises generating one or more automation scripts to execute the first set of actions on the application.

10. A computer program product for dynamic parametric modeling using learning reinforcement, the computer program product comprising a non-transitory computer-readable medium comprising code causing a first apparatus to:
electronically retrieve a first distributed impact simulation model, wherein the first distributed impact simulation model is generated using a first set of actions;
implement the first distributed impact simulation model on an application within a distributed environment, wherein implementing further comprises initiating the first set of actions on one or more application parameters;
initiate a reinforcement learning algorithm on the application, wherein initiating further comprises receiving a performance assessment output for the one or more application parameters based on at least initiating the first set of actions on the one or more application parameters;
initiate an optimization policy generation engine on the performance assessment output associated with the one or more application parameters to generate an optimization policy, wherein the optimization policy generation engine is configured to encode the performance assessment output into rewards and costs, wherein encoding further comprises assigning a cost to a first portion of the first set of actions and assigning a reward to a second portion of the first set of actions;
initiate an implementation of the optimization policy on the application, wherein initiating further comprises generating a second set of actions based on at least the optimization policy to maximize an aggregated reward calculated from the second portion of the first set of actions;
automatically generate a second distributed impact simulation model using the second set of actions to be implemented on the one or more application parameters; and
implement the second distributed impact simulation model on the application, wherein implementing further comprises initiating the second set of

Assignees

• Bank Of America

Inventors

• Krishnamoorthy Madhusudhanan
• Ameerhamza Ayesha Farha
• Venkatesan Abhiraam

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N7/02 Primary

  using fuzzy logic (computing arrangements based on biological models G06N3/00; computing arrangements using knowledge-based models G06N5/00) · CPC title

• G06N5/04

  Inference or reasoning models · CPC title

• G06N7/06

  Simulation on general purpose computers · CPC title

• G06N20/00 Primary

  Machine learning · CPC title