Kevin Greenman

@kevinpgreenman.bsky.social

🤖 AI for Science 🌟☧ Asst Prof. @CathInstTech | 🎓 Ph.D. @mitcheme.bsky.social‬ ’24 (@rgblabmit.bsky.social) | 💊💻 Ex: @msftresearch.bsky.social‬ & @elilillyandcompany.bsky.social & Chemprop maintainer | www.kevinpgreenman.com

We compared the calibration of various machine learning uncertainty estimation methods for protein engineering. No method excels across all scenarios, and uncertainty-based strategies for optimization often did not outperform methods without uncertainty.

Approach, datasets, and tasksMiscalibration area vs. root mean square error (RMSE) Active learningBayesian optimization