kakomu

A simple bot able to play game of go in Kotlin.

The general approach is inspired by the book "Deep Learning And The Game of Go". I introduced some optimizations and Kotlin idioms.

Currently it's using MonteCarlo Tree Search but hopefully it will also make use of DeepLearning.

The goals of Kakomu are:

• Enjoy and having fun

• Be able to experiment with MCTS, DeepLearning, etc

• Keep code simple and easy to understand and modify

• Get some hints on how to improve at Go

Not a goal:

• Make a top level AI player

• Create a framework for Go Game IA engine

BACKLOG (sort of)

SGF files load and save

GTF support to use a nicer GUI

DeepLearning training to guess best moves (esp. fuseki)

Multi threads support