This repository contains official implementaion of the paper "AdaFedProx: A Heterogeneity-aware Federated Deep Reinforcement Learning for Medical Image Classification". Steps to run the algorithms:

1. run Data_creation.py after changing the data distribution you want for all the three datsets 1) ISIC-2018 2) Pap Smear 3) Messidor dataset as described in the paper.
2. run all the algorithms including baseline and Proposed.
3. for proposed method first change the state for DQN agent as per the data distribution and system constarints as per described in the paper.