Intelligent Engineering Systems through Artificial Neural Networks, Volume 20
23 Competitive Coevolutionary Algorithm with Electric Charge Model for the Tsumego Game
Download citation file:
- Ris (Zotero)
- Reference Manager
This paper describes a Competitive Co-evolutionary Algorithm with electric charge model to solve Tsumego. The following three procedures achieve an effective evolutionary process. 1- The Go stones used in Tsumego are assumed to be electric charges. 2- In any specific arrangement of Go stones on the board in Tsumego, the electrical potential distribution of each intersection (where a stone is placed) on the board has a specific calculation. 3- Based on the electrical potential distribution (or variation of electric potential), intersections that do not reach a specific threshold level are excluded. (Intersections on the board are considered to be genes in the chromosomes of the evolution object. Therefore, the length of the chromosomes shortens according to number of excluded intersections.) Thus, the search domain and the calculation cost decreases. In addition, the numeric features calculated from the array of Go stones have been analyzed to develop more efficient methods.