How to use our code:
- The proposed MedlinePlus dataset can be found in ./environment/medlineplus.jsonCancel changes
- To run test the method on medlineplus,
- cd ./medlineplus_code
- CUDA_VISIBLE_DEVICES=0 python3 main.py -train -trail 1
- To test the method on symcat disease sets,
- cd ./symcat_code
- CUDA_VISIBLE_DEVICES=0 python3 main.py -train -train 1 -dataset 200 (300/400/common)
The citation for our paper is:
@misc{
efficientautodiag,
doi = {10.48550/ARXIV.2112.00733},
url = {https://arxiv.org/abs/2112.00733},
author = {Yuan, Hongyi and Yu, Sheng},
title = {Efficient Symptom Inquiring and Diagnosis via Adaptive Alignment of Reinforcement Learning and Classification},
publisher = {arXiv},
year = {2021},
copyright = {arXiv.org perpetual, non-exclusive license}
}