These files include all code to facilitate a LLM and supervised learning-guide NLP pipeline to identify social disconnection in the MIMIC-III Dataset.
Files should be run in order of:
- Lexicon Development
- Regular Expression Matching
- Classifiers (roberta, LLAMA-3, traditional ML)