David M. Blei

David M. Blei is the William B. Ransford Professor of Statistics and Computer Science at Columbia University. He is a leading researcher in machine learning, with a focus on probabilistic modeling and Bayesian statistics. His work bridges theoretical foundations with practical applications across various domains.

Professor Blei's research spans several key areas in modern machine learning:

• Development and analysis of probabilistic models for complex data
• Bayesian inference methods, particularly variational inference
• Topic modeling and mixed-membership models
• Causal inference and model criticism
• Deep generative models and representation learning
• Applications in natural language processing and recommendation systems

His recent publications demonstrate continued advancement in variational inference theory while exploring applications in deep learning and causality. Blei's work on posterior collapse in variational autoencoders, black box variational inference, and scalable recommendation systems has been particularly influential in the machine learning community.

Professor Blei mentors PhD students and postdoctoral researchers, fostering the next generation of machine learning researchers. He actively teaches graduate courses on probabilistic models, machine learning, and causal inference at Columbia University, including STCS 6701: Probabilistic Models and Machine Learning (scheduled for Fall 2025).

He leads a research group focused on probabilistic modeling, contributing significantly to both theoretical advancements and practical applications of statistical methods in artificial intelligence. The group is part of Columbia's thriving machine learning community, which spans multiple departments and research centers.