Open edgetrader opened 3 years ago
For recording purpose here
Code line 162
%%time model.fit(training_data[cols], training_data['direction'], epochs=50, verbose=False, validation_split=0.2, shuffle=False)
Should be to train model with normalised features
%%time model.fit(training_data_[cols], training_data['direction'], epochs=50, verbose=False, validation_split=0.2, shuffle=False)
Yes, I think the same.
For recording purpose here
Code line 162
Should be to train model with normalised features
Dr Yves' has replied here