Discovery of the Characteristics of the Cubic Othello Chessboard and Its Implementation of Visualization Expert System

EasyChair Preprint no. 12767, version 2

Abstract

The standard planar Othello game has 8(rows) * 8(columns) = 64(squares). Although this planar Othello game has been solved, this paper still proposes a new problem based on this foundation: Has the originally evenly matched situation between the first move and second move sides changed in the game of Othello played on a cubic board? Certainly, it is not very suitable for people to play against each other, because playing cubic Othello in a real-world environment is very difficult. Even if it is changed to the smallest cubic Othello board with 4(squares) * 6(faces) = 24(squares). Playing cubic Othello is very challenging because it requires understanding the characteristics of cubic Othello and making clear definitions of the ways of moving. Only in this way can we attempt to apply artificial intelligence (AI) techniques to cubic Othello and develop an expert system that can play chess on a cubic Othello board. Machine learning (ML) is a computer technique that uses a lot of input and output data to train software to understand correlations between the two. But before using ML techniques, we first used Monte Carlo simulations to predict the possible outcomes that would occur in cubic Othello. Monte Carlo simulation predicts that the winning rate of the first move (black) of cubic Othello is about 20%, while the winning rate of the second move (white) is about 80%. Clearly, in 4 (squares) * 6 (faces) cubic Othello, the second move has an advantage. Furthermore, the expert system proposed in this paper that is trained using the Web GPU on a personal computer can be executed on any contemporary browser. The training that originally took tens of days to complete using CPU memory on a personal computer can now be completed in tens of minutes in the Web GPU of a personal computer. This clearly shows that the significant benefits that can be achieved by effectively utilizing the GPU memory on a personal computer have surpassed the use of a large CPU memory computer.

Keyphrases: Artificial Intelligence, Cubic Chessboard, GPU memory, k-Nearest Neighbors Algorithm, machine learning, Othello Game

@Booklet{EasyChair:12767,
  author = {Jer Fong Chen and Fang Rong Hsu},
  title = {Discovery of the Characteristics of the Cubic Othello Chessboard and Its Implementation of Visualization Expert System},
  howpublished = {EasyChair Preprint no. 12767},

  year = {EasyChair, 2024}}