/clinical-ad

Abbreviation disambiguation (AD) in clinical notes using concept embedding model (Marta Skreta) and convolutional neural network baseline (Jacob Kelly).

Primary LanguageJupyter Notebook

clinical-ad

Code for "A conceptual route to unsupervised medical abbreviation disambiguation", final project for CSC2541: Machine Learning For Health

There are two Jupyter notebooks (and corresponding static HTML) reproducing the main results from our paper:

train_model/concept_embeddings/csc2541_finalProject_results.ipynb for the concept embedding model (Marta Skreta)

baselines/baselines.ipynb for the baseline models (Jacob Kelly)

Team

Marta Skreta, Jacob Kelly