Classifying the lifestyle status for Alzheimer's disease from clinical notes using deep learning with weak supervision.

The proposed approach leveraging weak supervision could significantly increase the sample size, which is required for training the deep learning models. By comparing with the traditional machine learning models, the study also demonstrates the high performance of BERT models for classifying lifestyle status for Alzheimer's disease in clinical notes.