Binary Classification on Pima Indians diabetes database (PIMA) (test)
0.84AUCLogistic Regression
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Logistic Regression2025.12 | 0.84 | — | |
| Random Forest with 100 treesnumber of trees=100, max depth=32025.12 | 0.84 | — | |
| Random Forest with 1000 treesnumber of trees=1000, max depth=32025.12 | 0.84 | — | |
| Soft Decision Treemax depth=32025.12 | 0.83 | — | |
| SM-SDTmax depth=3, short-term memory=true2025.12 | 0.83 | — | |
| Decision Treemax depth=32025.12 | 0.81 | — | |
| XGBoostmax depth=32025.12 | 0.81 | — | |
| CSP-BARTLink function=probit, Data augmentation scheme=Albert and Chib (1993), Covariate sharing/specification=Age and glucose in X1 and all 8 available covariates to the BART component2021.08 | — | 17.94 | |
| Hybrid modelLink function=probit, Data augmentation scheme=Albert and Chib (1993), Prior=isotropic prior from SSP-BART, Covariate sharing/specification=age and glucose in both components, Constraints=without the double moves and stringent checks on tree-structure validity used in CSP-BART2021.08 | — | 19.23 | |
| SSP-BARTLink function=probit, Data augmentation scheme=Albert and Chib (1993), Covariate sharing/specification=Age and glucose in X1 and the 6 remaining covariates in X22021.08 | — | 20.51 |