Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/rbroc/multidiagnosis-text
Feature-based and Transformer models for classification of mental disorders from text
https://github.com/rbroc/multidiagnosis-text
machine-learning nlp psychiatry text
Last synced: 28 days ago
JSON representation
Feature-based and Transformer models for classification of mental disorders from text
- Host: GitHub
- URL: https://github.com/rbroc/multidiagnosis-text
- Owner: rbroc
- License: mit
- Created: 2023-09-18T09:02:03.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-09-18T12:42:53.000Z (over 1 year ago)
- Last Synced: 2024-10-24T11:52:07.109Z (3 months ago)
- Topics: machine-learning, nlp, psychiatry, text
- Language: Jupyter Notebook
- Homepage: https://arxiv.org/abs/2301.06916
- Size: 20.3 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
### Multi-class diagnosis prediction using text
Code for training feature-based baselines and Transformer models for classification of mental disorders from text, both in a multiclass classification setting and in a binary classification setting.
- `train_baselines.py` contains code to train a variety of feature-based XGboost baselines, and baselines based on static vectors;
- `train_transformer.py` contains code to train transformer-based classifiers;
- `train_classifier.py` contains code to run baselines as classifications with a shallow neural network.
- `preproc.py` and `dataset.py` are dataset processing utils (data not shared due to confidentiality)The repository also includes additional utils and notebook for exploration of the results.
Models and their predictions are logged under `logs`. Aggregate model performance is reported in `logs/aggregates`, which also contains results from speech-based models.
The twin repository for audio-based models is here: https://github.com/HLasse/multidiagnosis-speechRelated publication (available [here](https://arxiv.org/abs/2301.06916)):
Hansen, L., Rocca, R., Simonsen, A., Parola, A., Bliksted, V., Ladegaard, N., ... & Fusaroli, R. (2023). Automated speech-and text-based classification of neuropsychiatric conditions in a multidiagnostic setting. arXiv preprint arXiv:2301.06916.