Table 2.
Main characteristics of the considered classification models
Model | Type | Feature importance API |
---|---|---|
Logistic regression | Generalized linear model | Yes |
Support vector machine | Supervised learning model constructing the separating manifold | Only for linear kernel |
XGBoost | Gradient-boosted decision tree ensemble | Yes |
CatBoost | Gradient-boosted decision tree ensemble | Yes |
LightGBM | Gradient-boosted decision tree ensemble | Yes |
TabNet | Deep neural network | Yes |
NODE | Gradient-boosted decision tree ensemble | No |