# tianshou
Tianshou(天授) is a reinforcement learning platform.

![alt text](https://github.com/sproblvem/tianshou/blob/master/docs/figures/tianshou_architecture.png "Architecture of tianshou")

## agent
    Examples

    Self-play Framework

## core

### Model
    DQN, Policy-Value Network of AlphaGo Zero, PPO-specific, TROP-specific

### Algorithm

#### Loss design
    Actor-Critic (Variations), DQN (Variations), DDPG, TRPO, PPO

#### Optimization method
    SGD, ADAM, TRPO, natural gradient, etc.

### Planning
    MCTS

## data
    Training style - Monte Carlo or Temporal Difference

    Reward Reshaping/ Advantage Estimation Function

    Importance weight

    Multithread Read/Write

## environment
    DQN repeat frames etc.

## simulator
    Go, Othello/Reversi, Warzone

## TODO
Search based method parallel.