Reinforcement Learning: An Introduction (2nd Edition) in C++
This is a C++ version of examples and demos from ShangtongZhang/reinforcement-learning-an-introduction. https://github.com/ShangtongZhang/reinforcement-learning-an-introduction
This project was inspired by the TicTacToe problem in the excellent book written by Prof Sutton. The Python example from Shangtong Zhang's great project demonstrates an excellent neat solution. However, the efficiency of Python limits the dimension and speed in running the TicTacToe. In addition, readers may be also interested in how to implement reinforcement learning in C++. This was the whole motivation of this project at the very beginning. As a result, the C++ version of TicTacToe can reach dimension=4 with reasonable speed.
This project is designed to be a supplement project of the Python version. For readers' convenience, all examples are written as close as possible to the Python version becides necessary enhancement.
Compile: Visual Studio 2015/2017 g++ 5.4.0 > g++ -std=c++11 -O *.cpp
Issues:
- no graphic output. used console output to replace graphic output in some examples.
- need fully test
Contribution: Welcome to contribute code/testing/bug_fix/ideas to this project.