Add more examples - Githubissues

ianozsvald / beyond_correlation

Exploratory code to see if we can learn about feature relationships in a DataFrame using machine learning

MIT License

55 stars 18 forks source link

Add more examples #3

Open ianozsvald opened 6 years ago

ianozsvald commented 6 years ago

Add new Notebooks with some interpretation, using the built-in sklearn datasets

[ ] diabetes
[ ] breast cancer
[ ] wine
[ ] text and numeric label e.g. [1, 2, 3] ['one', 'two', 'three'] many times - for a redundant column that should be identified
[ ] y=x^2 non-linear example which RF should do well on and Pearson should do poorly on

Hopefully also show corr results (especially if we get https://github.com/ianozsvald/discover_feature_relationships/issues/1 added)

Maybe also add Anscombe Quartet https://seaborn.pydata.org/examples/anscombes_quartet.html and use it in the README