Perform exploratory data analysis to the wine dataset from sklearn, we check correlations and predictive power of features through variable distribution graphics. Represent class distribution accross variables using 2D and 3D graphics and derive insights. Test PCA against TSNE, compare results and explanation. Testing the effect on a linear classifier of Random Oversampling vs SMOTE vs ADASYN in a target imbalanced engine failure dataset.
zetag33 / eda Goto Github PK
View Code? Open in Web Editor NEWExploratory data analysis, PCA/TSNE, Random Oversampling vs SMOTE vs ADASYN