18371270. SYSTEM AND METHOD FOR ACCELERATING BENDERS DECOMPOSITION VIA REINFORCEMENT LEARNING SURROGATE MODELS simplified abstract (JPMorgan Chase Bank, N.A.)

SYSTEM AND METHOD FOR ACCELERATING BENDERS DECOMPOSITION VIA REINFORCEMENT LEARNING SURROGATE MODELS

Organization Name

JPMorgan Chase Bank, N.A.

Inventor(s)

Kyle Mana of Renton WA (US)

Parisa Zehtabi of London (GB)

Hoa Lun Stephen Mak of Glasgow (GB)

Michael Cashmore of Stirlingshire (GB)

Daniele Magazzeni of London (GB)

Manuela Veloso of New York NY (US)

Daniel A Kirsner of Oak Park IL (US)

Ryan Eiben of Lewis Center OH (US)

Donald Stephens of Hollis NY (US)

Lei Carol Liang of Hamburg NJ (US)

Fernando Acero of London (GB)

SYSTEM AND METHOD FOR ACCELERATING BENDERS DECOMPOSITION VIA REINFORCEMENT LEARNING SURROGATE MODELS - A simplified explanation of the abstract

This abstract first appeared for US patent application 18371270 titled 'SYSTEM AND METHOD FOR ACCELERATING BENDERS DECOMPOSITION VIA REINFORCEMENT LEARNING SURROGATE MODELS

Simplified Explanation:

The patent application discusses methods and systems for accelerating decomposition through reinforcement learning. It involves using a processor to implement a decomposition algorithm for solving large linear programming problems with a block structure, inserting a reinforcement learning agent into the algorithm, and generating master problem decisions to replace NP-hard mixed-integer master problems.

• Reinforcement learning integrated into a decomposition algorithm
• Accelerating decomposition of large linear programming problems
• Replacement of NP-hard mixed-integer master problems with master problem decisions
• Use of a processor to implement the algorithm
• Special block structure in the linear programming problems

Potential Applications: The technology could be applied in various industries such as logistics, finance, and manufacturing where optimization of complex problems is required.

Problems Solved: The technology addresses the challenge of efficiently solving large linear programming problems with a special block structure by incorporating reinforcement learning.

Benefits: - Faster decomposition of complex linear programming problems - Improved efficiency in decision-making processes - Enhanced optimization capabilities in various industries

Commercial Applications: Optimizing supply chain management, financial portfolio management, and production planning processes using the accelerated decomposition technology.

Prior Art: Researchers can explore prior studies on reinforcement learning in optimization algorithms and decomposition methods to understand the evolution of this technology.

Frequently Updated Research: Stay updated on advancements in reinforcement learning algorithms and their applications in optimization to enhance the efficiency of decomposition processes.

Questions about Accelerating Decomposition via Reinforcement Learning: 1. How does reinforcement learning improve the efficiency of solving large linear programming problems? 2. What are the key differences between traditional decomposition methods and the approach involving reinforcement learning?

Original Abstract Submitted

Various methods, apparatuses/systems, and media for accelerating decomposition via reinforcement learning are disclosed. A processor implements a decomposition algorithm that allows a solution of a comparatively larger linear programming problems that have a special block structure; inserts a reinforcement learning agent within a framework of the decomposition algorithm; and generates, in response to inserting the reinforcement learning agent, master problem decisions in place of an NP-hard (nondeterministic polynomial time-hard) mixed-integer master problem (MIMP).