I wrote a Connect-6 player based on java during first year in university. And it's my first trial for AI designing. But limited to my ability at that moment, it's not well in use.
Bloody99 is an addition poker game modified from 99, which is well-known in Taiwan. Our team make this for our AI class project, trying for various game-solving algorithm like alpha-beta searching, monte-carlo method and heuristics. Also including GUI by PyGame.
I wrote this project for my favortite class "Theory of Computer Games" and get good result at peer contest, and get into Professor's lab for further research. Go is based on Monte-Carlo method. I further try statistics method to fortify it. After AlphaGo introduced in 2016/1/27, I also make use of neural network to get better pruning strategy.
Nonogram is a japanese puzzle that fill the board with color with row and column constrant. This is the first project in class "Theory of Computer Games", I first try heuristic to solve the game. But it's not assured to solve when there's more than one solution. So needs DFS search to help. This is also not perform well in performance.