CSIR-NCL ยท 2023
CancerXAI - Explainable Cancer Drug Recommendation
An explainable Streamlit application that lets oncology professionals compare machine learning models for cancer drug value recommendation.
What I built
Working with senior CSIR scientists, I built a web app that compares six selectable ML models and provides LIME explanations for each prediction. The interface makes model behavior easier to inspect while supporting practical drug-value analysis.
Recognition
- Reached 97% accuracy across the evaluated models.
- Co-authored two published papers on the project.
- Received a copyright grant for the CancerXAI system.