Removing Spurious Features Can Hurt Accuracy and Affect Groups Disproportionately

This course aims to educate learners on the potential negative impacts of removing spurious features on accuracy and its disproportionate effects on different groups. The course covers topics such as the introduction, accuracy drop, setup, data sets, results, other results, and conclusion. The teaching method involves presenting research findings in a concise 21-minute session. This course is intended for individuals interested in the intersection of feature selection, accuracy, and group fairness in computing and data analysis.