elgarsstone

Challenge 2b – Data Quality, Fairness & SQL Sampling (AIF360)

This notebook evaluates data quality and fairness in a socio-economic dataset using the IBM AI Fairness 360 (AIF360) toolkit.

It includes: • Data quality inspection and descriptive statistics
• Bias detection using Statistical Parity Difference and Disparate Impact
• Bias mitigation with the Reweighing preprocessing algorithm
• Balanced sampling implemented with SQL (ROW_NUMBER partitioning)

elgarsstone

Challenge 2a – DEI-aware Data Preparation

This notebook introduces Diversity, Equity, and Inclusion (DEI) concepts in data-driven experiments. It focuses on identifying potential sources of bias in human-related datasets and highlights how data preparation choices can influence fairness outcomes.

The notebook serves as an introductory step toward fairness-aware data analysis, preparing the ground for bias detection, mitigation, and fair modelling explored in later challenges.