Daniel Zeiberg is a postdoctoral fellow at the Northeastern Institute for Experiential AI. His current research focuses on developing machine learning methods for protein function and variant effect prediction. Daniel’s academic interest is in uncovering biological insights using cutting-edge deep learning tools, including protein structure predictors and large language models.

Daniel’s past research focused on developing ways to train classifiers to predict calibrated probabilities from data that violate standard machine learning assumptions. He has addressed forms of statistical bias such as positive-unlabeled learning, covariate shift, and label shift. He has applied this expertise to genomic medicine, developing a novel method for calibrating functional assays as evidence for clinical variant classification. Before joining the Institute for Experiential AI, he received his doctorate from the Khoury College of Computer Sciences at Northeastern University.

‍