MIT EECS | Angle Undergraduate Research and Innovation Scholar
Imitating Human Chess Performance Using Deep Neural Network
- Artificial Intelligence and Machine Learning
As artificially intelligent chess programs continue achieving higher and higher levels of superhuman performance, the ability of humans to learn from these programs is becoming increasingly difficult. Interpreting the reasons behind why a computer program makes a specific move in a specific position is often limited by one’s understanding of the program itself. This project focuses on using generative adversarial networks to imitate human chess play rather than predict the move that maximizes the likelihood of winning.