search

justinetaylor / mids-w207-final-project

Clear Cut Solution - https://www.kaggle.com/c/forest-cover-type-prediction
1 stars 1 forks source link

Yang branch #11

Closed lateflare closed 4 years ago

lateflare commented 4 years ago

Updates scaling.

justinetaylor commented 4 years ago

We need to scale the dev data based on the training data mean and standard deviation. I think scaler.fit_and_transform(devdata) will calculate use the dev data mean/standard deviation. Can you update this to standardize the dev data using scaler.mean and scaler.var_**(1/2) ?