AI, Data Science & Cloud Certificates from Google, IBM & Meta — 40% Off
One plan covers every Professional Certificate on Coursera. 40% off your first 3 months — limited time.
Unlock All Certificates
Explore a 20-minute ACM conference talk examining the implications of multimodal CLIP filtering in DataComp through an empirical case study. Delve into research findings presented by Rachel Hong, William Agnew, Tadayoshi Kohno, and Jamie Morgenstern as they analyze the filtering mechanisms and their impact on data inclusion and exclusion patterns. Learn about the critical aspects of data curation in machine learning datasets and understand how CLIP-based filtering decisions affect the composition and representation within DataComp.
