Chalk Talk: Getting Started with Probabilistic Topic Models

Kim, Yoonsang. (2015, Nov. 10) Statistical Getting Started with Probabilistic Topic Models [PDF of PowerPoint slides]. Institute for Health Research and Policy, University of Illinois at Chicago.

A topic model is an unsupervised machine learning technique. Probabilistic topic modeling is used to discover the themes in a large unstructured collection of documents by analyzing the words in documents. The talk focused on the simple topic model, latent Dirichlet allocation. 

View Dr. Kim's presentation. (PDF) 

Health researchers at all levels, including postdoctoral investigators and doctoral students designing their own studies, are invited.

MRC Chalk Talks are informal educational seminars about current methodological issues hosted by the IHRP Methodology Research Core and held on the second Tuesday of the month at 1 p.m.