Solving Atari Games Using Fractals And Entropy

TL;DR

This paper presents Fractal Monte Carlo, a thermodynamics-inspired MCTS-based algorithm that significantly improves efficiency in Atari game environments by enabling more intelligent agent actions with controllable behavior.

Contribution

Introduces Fractal Monte Carlo, a novel thermodynamics-based MCTS approach that enhances efficiency and control in agent decision-making for Atari games.

Findings

FMC outperforms traditional MCTS by several orders of magnitude in efficiency.

FMC effectively handles both continuous and discrete environments.

The approach provides detailed control over agent behavior.

Abstract

In this paper, we introduce a novel MCTS based approach that is derived from the laws of the thermodynamics. The algorithm coined Fractal Monte Carlo (FMC), allows us to create an agent that takes intelligent actions in both continuous and discrete environments while providing control over every aspect of the agent behavior. Results show that FMC is several orders of magnitude more efficient than similar techniques, such as MCTS, in the Atari games tested.