Responsible Prediction Making of COVID-19 Mortality

Paper

Baniecki, H., and Biecek, P. Responsible Prediction Making of COVID-19 Mortality (Student Abstract)
AAAI Conference on Artificial Intelligence (AAAI) 2021 URL: https://ojs.aaai.org/index.php/AAAI/article/view/17874

Abstract

For high-stakes prediction making, the Responsible Artificial Intelligence (RAI) is more important than ever. It builds upon Explainable Artificial Intelligence (XAI) to advance the efforts in providing fairness, model explainability, and accountability to the AI systems. During the literature review of COVID-19 related prognosis and diagnosis, we found out that most of the predictive models are not faithful to the RAI principles, which can lead to biassed results and wrong reasoning. To solve this problem, we show how novel XAI techniques boost transparency, reproducibility and quality of models.

modelStudio.gif

Dashboard rai-covid

contribution0.40.60.81.0LDHhs-CRPAgeLymphocytes(%)Shapley ValuesDOne minus AUC0.000.020.040.060.080.100.12LDHLymphocytes(%)hs-CRPAgeFeature ImportanceDLDHPartial Dependence-15001,0001,5001,8670.20.30.40.50.60.70.8average predictionDLDHAverage Target vs Feature-15001,0001,5001,8670.00.20.40.60.81.0average targetInteractive Studio for Xgboost COVID19 ModelSite built with modelStudio v2.1.1 on 2021-05-30 17:20:32recall: 1 | precision: 0.994 | f1: 0.997 | accuracy: 0.997 | auc: 1XXXX

Created using modelStudio: https://github.com/ModelOriented/modelStudio

Preliminary code for this work is available: https://github.com/hbaniecki/Pre-Surv-COVID-19

Citation

To cite this work, use the following BibTeX entry:

@article{rai-covid,
  title={{Responsible Prediction Making of COVID-19 Mortality (Student Abstract)}},
  author={Baniecki, Hubert and Biecek, Przemyslaw},
  journal={AAAI Conference on Artificial Intelligence (AAAI)},
  year={2021},
  url={https://ojs.aaai.org/index.php/AAAI/article/view/17874}
}