Hybrid computer approach to train a machine learning system

TL;DR

This paper presents a hybrid digital-analog computer system for training machine learning models, demonstrating its effectiveness with a reinforcement learning task involving an inverted pendulum simulation.

Contribution

It introduces a novel hybrid computer setup that combines digital and analog computing to improve environment simulation for reinforcement learning.

Findings

Successful training of a reinforcement learning system using the hybrid setup

Effective simulation of the environment with an analog computer

Potential for improved training efficiency in machine learning

Abstract

This book chapter describes a novel approach to training machine learning systems by means of a hybrid computer setup i.e. a digital computer tightly coupled with an analog computer. As an example a reinforcement learning system is trained to balance an inverted pendulum which is simulated on an analog computer, thus demonstrating a solution to the major challenge of adequately simulating the environment for reinforcement learning.