Evaluating Unlearning and Memorization via Compression

Evaluating Unlearning and Memorization via Compression

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Learn about innovative approaches to evaluating machine unlearning and memorization through compression techniques in this 55-minute research presentation by CMU researcher Zhili Feng at the Center for Language & Speech Processing. Explore cutting-edge methodologies for measuring how well machine learning models can forget specific data points while maintaining overall performance, and discover new perspectives on quantifying memorization behaviors through the lens of data compression principles.