D
David M. Blei
Researcher at Columbia University
Publications - 399
Citations - 122384
David M. Blei is an academic researcher from Columbia University. The author has contributed to research in topics: Inference & Topic model. The author has an hindex of 98, co-authored 378 publications receiving 111547 citations. Previous affiliations of David M. Blei include Columbia University Medical Center & Hewlett-Packard.
Papers
More filters
Journal ArticleDOI
Distance Dependent Infinite Latent Feature Models
TL;DR: In this article, a distance dependent Indian buffet process (dd-IBP) is proposed to model non-exchangeable data, which relies on distances defined between data points, biasing nearby data to share more features.
Posted Content
Using Embeddings to Correct for Unobserved Confounding in Networks
TL;DR: The main idea is to reduce the causal estimation problem to a semi-supervised prediction of both the treatments and outcomes and networks admit high-quality embedding models that can be used for this semi- supervised prediction.
Posted Content
SHOPPER: A Probabilistic Model of Consumer Choice with Substitutes and Complements
TL;DR: ShopPER as mentioned in this paper uses interpretable components to model the forces that drive how a customer chooses products; in particular, they designed SHOPPER to capture how items interact with other items, and developed an efficient posterior inference algorithm to estimate these forces from large-scale data, and analyzed a large dataset from a major chain grocery store.
Proceedings Article
A Bayesian Analysis of Dynamics in Free Recall
Richard Socher,Samuel J. Gershman,Per B. Sederberg,Kenneth A. Norman,Adler J. Perotte,David M. Blei +5 more
TL;DR: A probabilistic model of human memory performance in free recall experiments is developed, conceptualizing memory retrieval as a dynamic latent variable model and using Bayesian inference to represent uncertainty and reason about the cognitive processes underlying memory.
Proceedings Article
Variational Inference for Adaptor Grammars
TL;DR: A variational inference algorithm for adaptor grammars is described, providing an alternative to Markov chain Monte Carlo methods, and a significant speed-up is shown when parallelizing the algorithm.