Classifying Free-Text Diagnostics
In this workshow we will cover the entire pipeline to train a classifier to predict the specialty where a patient should be referred based on the unstructured text of his diagnostic suspicion.
You will learn:
- Text preprocessing
- Train word embeddings
- Use word embeddings
- Compose sentence embeddings
- Train a machine learning classifier from word embeddings as features
To use the code of the workshop you could clone this repository, install the requirements and launch the main jupyter notebook. Or you can launch the notebook ready-to-use from the binder virtual machine.