Most of the tasks is done in Orange apart from two tasks:
1. Obtaining skewness and kurtosis values.
2. Split the dataset into training set and test set using random_state=309
The two part above is done in python Jupyter notebook: ASS2.ipynb
note: Data file path need to be adjusted if you run the jupter notebook on a different machine.


The orange workflow for Part 2 and Part 3 has PCA connected with the test and score widget for evaluation. 
If you need to evaluate the Rank method, simply connect Rank with the test and score widget which will replace the PCA connection automatically.