Approximation of Convex Envelope Using Reinforcement Learning

2023/11/24 by Vivek S. Borkar, Borkar, Vivek S., Adit Akarsh +1
Computer Science · Decision Sciences · #Auction Theory and Applications #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Reinforcement Learning in Robotics #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2311.14421

openalex publication_date 2023/11/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Oberman gave a stochastic control formulation of the problem of estimating the convex envelope of a non-convex function. Based on this, we develop a reinforcement learning scheme to approximate the convex envelope, using a variant of Q-learning for controlled optimal stopping. It shows very promising results on a standard library of test problems.

Related

• Sufficient Exploration for Convex Q-learning
• Convex Q-Learning, Part 1: Deterministic Optimal Control
• Convex Q Learning in a Stochastic Environment: Extended Version
• Difference of Convex Functions Programming Applied to Control with\n Expert Data
• Reinforcement Learning via Fenchel-Rockafellar Duality