Implementing Online Reinforcement Learning with Temporal Neural Networks

TL;DR

This paper introduces a novel Temporal Neural Network architecture designed for efficient online reinforcement learning, combining unsupervised clustering and reinforcement learning with biologically plausible learning rules, demonstrated through a cart-pole simulation.

Contribution

It presents a new TNN-based system integrating unsupervised clustering and reinforcement learning with biologically inspired rules for online learning.

Findings

Successful simulation of cart-pole balancing using the proposed TNN system

Demonstration of online unsupervised clustering and reinforcement learning integration

Validation of biologically plausible learning rules in a reinforcement learning context

Abstract

A Temporal Neural Network (TNN) architecture for implementing efficient online reinforcement learning is proposed and studied via simulation. The proposed T-learning system is composed of a frontend TNN that implements online unsupervised clustering and a backend TNN that implements online reinforcement learning. The reinforcement learning paradigm employs biologically plausible neo-Hebbian three-factor learning rules. As a working example, a prototype implementation of the cart-pole problem (balancing an inverted pendulum) is studied via simulation.