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

why the rewards are different between train and predict? (ES) #120

Open tiger55cn opened 1 year ago

tiger55cn commented 1 year ago

For example, if you train the data 100 times, at the end, the reward is 150%. With exactly the same trained agent, the reward returned by buy() is 130%. The data is the same, the weight is the same, but the rewards are different. Is that a bug?