TopicModeling_LDA

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.

See the full implementation and results in LDA.ipynb.

Learned Topics

Collapsed Gibbs sampling implementation borrowed & adapted from CS179 course material by Prof. Alexander Ihler.

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
assets		assets
.DS_Store		.DS_Store
.gitattributes		.gitattributes
.gitignore		.gitignore
LDA.ipynb		LDA.ipynb
README.md		README.md
TopicModeling_LDA.ipynb		TopicModeling_LDA.ipynb
requirements.txt		requirements.txt