Run generate_friedman.ipynb to have friedman data of size 6k
Then, go through the Friedman_test.ipynb
After done the splitting, move to MASATML (takes around 10min per test )
Generate_friedman.ipynb: Run the notebook and it can generate friedman data with 5 features and N samples. Very simple.
Friedman_test.ipynb: The notebook used to split data based on mahalanobis distance. It also used to visualize the final result produced by MASTML. You can also apply on other data sets.