search

huseinzol05 / Stock-Prediction-Models

Gathers machine learning and deep learning models for Stock forecasting including trading bots and simulations
Apache License 2.0
7.66k stars 2.74k forks source link

Q learning not implemented properly #90

Open shanyaanand opened 3 years ago

shanyaanand commented 3 years ago

Hi, I really appreciate your work however I have noticed lot of issue/problems.

  1. There is exploration vs exploitation in action in q learning, you are just using greedy action policy which is not optimal solution
  2. The reason why deep qlearning works in because of replay memory and target and estimator network but in the code you have combined both of them and there is not replay memory.
  3. you are not optimising Bellmann equation.