Skip to content

Latest commit

 

History

History
6 lines (5 loc) · 778 Bytes

README.md

File metadata and controls

6 lines (5 loc) · 778 Bytes

Time Series Analysis

Repository for Jupyter Notebooks and Python scripts that implement time series analysis for BIDMC Digital Psychiatry Studies.

Paper Title Relevant File Names Description
Predicting PANSS and SFS Scores with LSTM Models: An Exploratory Time Series Analysis of Digital Phenotyping Data in Schizophrenia LSTM_RNN_PANSS_SFS_Score_Prediction.ipynb This Jupyter notebook loads in forward filled data from the SHARP dataset containing active, passive, and structured survey data. It preprocesses the data for time series analysis, creates RNN and LSTM models, trains the models on a training subset of the data, and evaluates performance using KFold cross-validation on a test subset of data.