Hybrid fighting game AI using a genetic algorithm and Monte Carlo tree search

2018 ◽  
Author(s):  
Man-Je Kim ◽  
Chang Wook Ahn
Keyword(s):  
Genetic Algorithm ◽  
Monte Carlo ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Game Ai

scholarly journals A graph-based genetic algorithm and generative model/Monte Carlo tree search for the exploration of chemical space

2019 ◽  
Vol 10 (12) ◽  
pp. 3567-3572 ◽  
Author(s):  
Jan H. Jensen
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Monte Carlo ◽  
Chemical Space ◽  
Generative Model ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Synthetic Accessibility ◽  
Better Than

This paper presents a comparison of a graph-based genetic algorithm (GB-GA) and machine learning (ML) results for the optimization of log P values with a constraint for synthetic accessibility and shows that the GA is as good as or better than the ML approaches for this particular property.

scholarly journals Graph-based Genetic Algorithm and Generative Model/Monte Carlo Tree Search for the Exploration of Chemical Space

2018 ◽  
Author(s):  
Jan H. Jensen
Keyword(s):  
Genetic Algorithm ◽  
Monte Carlo ◽  
Chemical Space ◽  
Generative Models ◽  
Generative Model ◽  
Small Data ◽  
Data Sets ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Small Data Sets

This paper presents a comparison of a graph-based genetic algorithm (GB-GA) and machine learning (ML) results for the optimisation of logP values with a constraint for synthetic accessibility and shows that GA is as good or better than the ML approaches for this particular property. The molecules found by GB-GA bear little resemblance to the molecules used to construct the initial mating pool, indicating that the GB-GA approach can traverse a relatively large distance in chemical space using relatively few (50) generations. The paper also introduces a new non-ML graph-based generative model (GB-GM) that can be parameterized using very small data sets and combined with a Monte Carlo tree search (MCTS) algorithm. The results are comparable to previously published results (Sci. Technol. Adv. Mater. 2017, 18, 972-976) using a recurrent neural network (RNN) generative model, while the GB-GM-based method is orders of magnitude faster. The MCTS results seem more dependent on the composition of the training set than the GA approach for this particular property. Our results suggest that the performance of new ML-based generative models should be compared to more traditional, and often simpler, approaches such a GA.

Application of Monte-Carlo tree search in a fighting game AI

2016 ◽  
Author(s):  
Shubu Yoshida ◽  
Makoto Ishihara ◽  
Taichi Miyazaki ◽  
Yuto Nakagawa ◽  
Tomohiro Harada ◽  
...  
Keyword(s):  
Monte Carlo ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Game Ai

scholarly journals Graph-based Genetic Algorithm and Generative Model/Monte Carlo Tree Search for the Exploration of Chemical Space

2019 ◽  
Author(s):  
Jan H. Jensen
Keyword(s):  
Genetic Algorithm ◽  
Monte Carlo ◽  
Chemical Space ◽  
Generative Models ◽  
Generative Model ◽  
Small Data ◽  
Data Sets ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Small Data Sets

This paper presents a comparison of a graph-based genetic algorithm (GB-GA) and machine learning (ML) results for the optimisation of logP values with a constraint for synthetic accessibility and shows that GA is as good or better than the ML approaches for this particular property. The molecules found by GB-GA bear little resemblance to the molecules used to construct the initial mating pool, indicating that the GB-GA approach can traverse a relatively large distance in chemical space using relatively few (50) generations. The paper also introduces a new non-ML graph-based generative model (GB-GM) that can be parameterized using very small data sets and combined with a Monte Carlo tree search (MCTS) algorithm. The results are comparable to previously published results (Sci. Technol. Adv. Mater. 2017, 18, 972-976) using a recurrent neural network (RNN) generative model, while the GB-GM-based method is orders of magnitude faster. The MCTS results seem more dependent on the composition of the training set than the GA approach for this particular property. Our results suggest that the performance of new ML-based generative models should be compared to more traditional, and often simpler, approaches such a GA.

scholarly journals An Analysis of Heuristics for a Mathematically Incomplete Variant of TicTacToe

2021 ◽  
Author(s):  
Feyaz Baker ◽  
Arunava Mukhoti ◽  
B. R. Chandavarkar
Keyword(s):  
Monte Carlo ◽  
Mathematical Analysis ◽  
Optimal Strategy ◽  
Optimal Strategies ◽  
Public Domain ◽  
Tree Search ◽  
A Posteriori ◽  
Monte Carlo Tree Search ◽  
The Public ◽  
Game Ai

We attempt to produce a game-winning heuristic for the mathematically incomplete game Ultimate Tic Tac Toe (UTT). There are several game AI that use Monte Carlo Tree Search to decide moves, however, heuristics offer a faster and computationally cheaper alternative. The mathematical analysis of UTT has not been actively pursued, so we attempt to prove a posteriori. We have decided on a few strategies for playing, and assign different strategies to each player. We play several automated games of UTT, and statistically analyse which games end quickest, and use that data to find optimal strategies for playing. This can be used to produce game heuristics for more complicated games, and produce insight about strategies. The first objective is to specify a framework that can compare heuristics for UTT, and decide an optimal strategy for both players. The second objective is to test the framework with a large amount of data, and produce demonstrable results for UTT. Lastly, to aid further research in this topic, we release our dataset into the public domain.

scholarly journals Graph-based Genetic Algorithm and Generative Model/Monte Carlo Tree Search for the Exploration of Chemical Space

2019 ◽  
Author(s):  
Jan H. Jensen
Keyword(s):  
Genetic Algorithm ◽  
Monte Carlo ◽  
Chemical Space ◽  
Generative Models ◽  
Generative Model ◽  
Small Data ◽  
Data Sets ◽  
Tree Search ◽  
Monte Carlo Tree Search ◽  
Small Data Sets

This paper presents a comparison of a graph-based genetic algorithm (GB-GA) and machine learning (ML) results for the optimisation of logP values with a constraint for synthetic accessibility and shows that GA is as good or better than the ML approaches for this particular property. The molecules found by GB-GA bear little resemblance to the molecules used to construct the initial mating pool, indicating that the GB-GA approach can traverse a relatively large distance in chemical space using relatively few (50) generations. The paper also introduces a new non-ML graph-based generative model (GB-GM) that can be parameterized using very small data sets and combined with a Monte Carlo tree search (MCTS) algorithm. The results are comparable to previously published results (Sci. Technol. Adv. Mater. 2017, 18, 972-976) using a recurrent neural network (RNN) generative model, while the GB-GM-based method is orders of magnitude faster. The MCTS results seem more dependent on the composition of the training set than the GA approach for this particular property. Our results suggest that the performance of new ML-based generative models should be compared to more traditional, and often simpler, approaches such a GA.

Sign in / Sign up

Export Citation Format

Share Document