Add new linters (should support both data and data+code contexts): - high dimensionality (#59) - highly correlated/colinear features (#41) - insufficient data samples (#59) - data sparsity - high cardinality (#59) - data leakage, AST-based (i.e. preprocessing before data split, incorrect data split) - wrong distribution (w. for specific code) - lack of explainability / interpretability / transparency
Add new linters (should support both data and data+code contexts):