https://github.com/lucastheis/logistic_lda

Basic tensorflow implementation of logistic latent Dirichlet allocation
https://github.com/lucastheis/logistic_lda

classification lda machine-learning tensorflow topic-modeling

Last synced: about 2 months ago
JSON representation

Basic tensorflow implementation of logistic latent Dirichlet allocation

• Host: GitHub
• URL: https://github.com/lucastheis/logistic_lda
• Owner: lucastheis
• License: other
• Created: 2019-06-28T08:40:14.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2019-10-17T16:46:40.000Z (almost 5 years ago)
• Last Synced: 2024-04-08T03:01:55.376Z (6 months ago)
• Topics: classification, lda, machine-learning, tensorflow, topic-modeling
• Language: Python
• Homepage:
• Size: 48.8 KB
• Stars: 18
• Watchers: 2
• Forks: 5
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-topic-models - Logistic LDA - Tensorflow implementation of Discriminative Topic Modeling with Logistic LDA [:page_facing_up:](https://proceedings.neurips.cc/paper/2019/file/54ebdfbbfe6c31c39aaba9a1ee83860a-Paper.pdf) (Models / Miscellaneous topic models)

README

        
Logistic LDA

============

This package provides basic implementations of _logistic latent Dirichlet allocation_. It can be

used to discover topics in data containing groups of thematically related items, using either

labeled data or unlabeled data.

If you want to reproduce experiments of our paper, start here instead instead: [:octocat: logistic-lda/experiments](https://github.com/lucastheis/logistic_lda/tree/experiments)

Requirements

------------

* tensorflow == 1.13.2

* numpy >= 1.16.4

The code was tested with the versions above, but older versions might also work.

Getting started

---------------

To get started, download a version of the 20-Newsgroups dataset in TFRecord format:

	./scripts/download_news20.sh

Once downloaded, training can be started with:

	./scripts/train_news20.sh

To use your own dataset, take a look at `./logistic_lda/data.py` for a description of the data

format expected by the training script. Alternatively, modify the training script to use datasets

not stored as TFRecords.

After training has finished, compute predictions on another dataset and evaluate accuracy:

	./scripts/evaluate_news20.sh

The results of the evaluation can be found in `./models/news20/`.

Reference

---------

I. Korshunova, H. Xiong, M. Fedoryszak, L. Theis  

*Discriminative Topic Modeling with Logistic LDA*  

Advances in Neural Information Processing Systems 33, 2019