Research2025-03
A cross dataset meta-model for hepatitis C detection using multi-dimensional pre-clustering
Multi-dimensional pre-clusteringDeep LearningExplainability
Overview
Built a cross dataset meta-model for hepatitis C detection using multi-dimensional pre-clustering. The model uses a novel multi-dimensional pre-clustering approach in a cross data-set meta-model to detect hepatitis C in the patients non-invasively. We also verify the model's explainability through SHAP and LIME values.
Key Results
Novel multi-dimensional pre-clustering approach in a cross data-set meta-model to detect hepatitis C in the patients non-invasively.
Verified the model's explainability through SHAP and LIME values for both the individual models and the cross data-set meta-model.
Achieved 94.82% accuracy on the test set using novel method.
Accepted in Scientific Reports.
Tech Stack
TensorFlowKerasNumPyPandasScikit-learnSHAPLIME