Skip to content

Latest commit

 

History

History
13 lines (5 loc) · 346 Bytes

File metadata and controls

13 lines (5 loc) · 346 Bytes

PubMed TopicModeling

Latent Dirichlet Allocation (LDA) applied to 2,000 PubMed abstracts using collapsed Gibbs sampling to discover 20 latent biomedical topics.

output quality is evaluated using topic coherence metrics (c_v & c_npmi), indicating interpretable biomedical topics.

Learned Topics

Word Clouds