Solving Atari Games Using Fractals And Entropy
This addresses the problem of inefficient decision-making in reinforcement learning for game environments, offering a potentially transformative method for AI agents.
The paper tackles the problem of creating efficient agents for Atari games by introducing Fractal Monte Carlo (FMC), a novel MCTS-based approach derived from thermodynamics, which results in several orders of magnitude greater efficiency compared to similar techniques like MCTS.
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.