Multi-timescale Nexting in a Reinforcement Learning Robot

TL;DR

This paper demonstrates a robot's ability to learn and predict thousands of environmental features across multiple timescales in real time using reinforcement learning techniques, achieving high accuracy and efficiency.

Contribution

It introduces a scalable, real-time multi-timescale prediction system for robots using temporal-difference learning with linear function approximation.

Findings

Learns 2000 predictions dependent on 6000 features at over 10Hz.

Most learning completes within 30 minutes.

Predictions are comparable to optimal offline solutions.

Abstract

The term "nexting" has been used by psychologists to refer to the propensity of people and many other animals to continually predict what will happen next in an immediate, local, and personal sense. The ability to "next" constitutes a basic kind of awareness and knowledge of one's environment. In this paper we present results with a robot that learns to next in real time, predicting thousands of features of the world's state, including all sensory inputs, at timescales from 0.1 to 8 seconds. This was achieved by treating each state feature as a reward-like target and applying temporal-difference methods to learn a corresponding value function with a discount rate corresponding to the timescale. We show that two thousand predictions, each dependent on six thousand state features, can be learned and updated online at better than 10Hz on a laptop computer, using the standard TD(lambda) algorithm with linear function approximation. We show that this approach is efficient enough to be practical, with most of the learning complete within 30 minutes. We also show that a single tile-coded feature representation suffices to accurately predict many different signals at a significant range of timescales. Finally, we show that the accuracy of our learned predictions compares favorably with the optimal off-line solution.