Bandit-Based Genetic Programming

نویسندگان

Jean-Baptiste Hoock

Olivier Teytaud

چکیده

We consider the validation of randomly generated patterns in a Monte-Carlo Tree Search program. Our bandit-based genetic programming (BGP) algorithm, with proved mathematical properties, outperformed a highly optimized handcrafted module of a well-known computer-Go program with several world records in the game of Go.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Bedload transport predictions based on field measurement data by combination of artificial neural network and genetic programming

Bedload transport is an essential component of river dynamics and estimation of its rate is important to many aspects of river management. In this study, measured bedload by Helley- Smith sampler was used to estimate the bedload transport of Kurau River in Malaysia. An artificial neural network, genetic programming and a combination of genetic programming and a neural network were used to estim...

متن کامل

Four proofs of Gittins' multiarmed bandit theorem

We study four proofs that the Gittins index priority rule is optimal for alternative bandit processes. These include Gittins’ original exchange argument, Weber’s prevailing charge argument, Whittle’s Lagrangian dual approach, and Bertsimas and Niño-Mora’s proof based on the achievable region approach and generalized conservation laws. We extend the achievable region proof to infinite countable ...

متن کامل

A Genetic Programming-based Scheme for Solving Fuzzy Differential Equations

This paper deals with a new approach for solving fuzzy differential equations based on genetic programming. This method produces some trial solutions and seeks the best of them. If the solution cannot be expressed in a closed analytical form then our method produces an approximation with a controlled level of accuracy. Furthermore, the numerical results reveal the potential of the proposed appr...

متن کامل

On 2-armed Gaussian Bandits and Optimization

We explore the 2-armed bandit with Gaussian payoos as a theoretical model for optimization. We formulate the problem from a Bayesian perspective, and provide the optimal strategy for both 1 and 2 pulls. We present regions of parameter space where a greedy strategy is provably optimal. We also compare the greedy and optimal strategies to a genetic-algorithm-based strategy. In doing so we correct...

متن کامل