Reinforcement learning models for warfarin dose online estimation.
This repository currently implements the following warfarin dose estimation models.
Assign 5mg/day to each patient.
See fixed_dose.py.
Estimate the warfarin dose by a linear combination of patient characteristics (height, weight, race, etc.).
See clinical_dosing.py.
LinUCB algorithm based on feature vector of length 2344 for each patient.
See lin_ucb.py.
Lasso bandit algorithm based on feature vector of length 2344 for each patient.
See lasso_bandit.py.
Reference: H. Bastani and M. Bayati. Online decision-making with high-dimensional covariates. 2015.