wjaskowski/TCIAIG-2015-Co-CMA-ES
Coevolutionary CMA-ES for Knowledge-Free Learning of Game Position Evaluation
Discovered public repositories for wjaskowski in the GitHub catalog.
Coevolutionary CMA-ES for Knowledge-Free Learning of Game Position Evaluation
Code repository for GECCO 2015 paper: "N-Tuple Network for Knowledge-Free Reinforcement Learning in High Dimensions: A Case Study in SZ-Tetris"
Game 2048 GECCO 2015 Competition
Code for Artificial and Computational Intelligence Methods Course at Poznan University of Technology, Poland
This repository contains the solution submitted by team J12 for the EURO/ROADEF 2012 Challange Machine Reassignment Problem proposed by Google
Performance profiles generation for Weighted Piece Counter (WPC) strategies of Othello