Table 5 Comparison of related research.
From: Enhanced cervical cancer diagnosis using a novel Bayesian fusion ensemble method with explainable AI
References | Applied feature selection | Best model found | Performance | XAI and Deployment |
|---|---|---|---|---|
Suvanasuthi et al.11 | PCA | ML-miRNA | ACC: 90.9% | NO |
Mathivanan et al.12 | NO | ML + ResNet-152 | ACC: 98.08% | NO |
Mohi Uddin et al.13 | PCA and XGBoost | ROS + XGBoost + PCA + Hard Voting | ACC: 99.19% | NO |
Shakil et al.14 | Chi-square, LASSO, and SHAP (XAI) | DT + Chi-square | ACC: 96.60% | SHAP only |
Lleberi et al.15 | PSO | RF + PSO | ACC: 98.00% | NO |
Glucina et al.19 | NO | MLP + KNN + Random Oversampling | ACC: 95.00% | NO |
Tanimu et al.25 | RFE and LASSO | DT | ACC: 98.72% | NO |
Jahan et al.28 | Chi-square and SelectBest | MLP | ACC: 98.10% | NO |
Our proposed model | Boruta | ROS + ICA + PCA + Bayesian Fusion | ACC: 99.88% | SHAP, LIME, Web Application (real-time) |
