cleanlabintermediate30 min
Cleanlab + data quality assessment
Automated detection of label errors, outliers, duplicates, and other data issues in ML datasets for AI agents handling real-world messy data.
Prerequisites
- →Python 3.8+
- →Install cleanlab[datalab] via pip
- →scikit-learn
- →numpy
- →pandas
- →Trained ML model for pred_probs/features
- →Labeled classification dataset in dict/DataFrame/HuggingFace format.
Tools used
Related Concepts
Browse full Lexicon →