What do "explained_var_old" and "explained_var_new" imply?

junxiaosong / AlphaZero_Gomoku

An implementation of the AlphaZero algorithm for Gomoku (also called Gobang or Five in a Row)

MIT License

3.29k stars 964 forks source link

What do "explained_var_old" and "explained_var_new" imply? #85

Closed bacTlink closed 5 years ago

bacTlink commented 5 years ago

I read your codes and I found two variables "explained_var_old" and "explained_var_new". Obviously they are for observing the convergence. But I can not figure out what they imply. Is there any relevant theory?

bacTlink commented 5 years ago

It's relative entropy.