Methods · Tabular
Classic / statistical fits via scikit-learn. Boosting prefers XGBoost → LightGBM → CatBoost → sklearn when installed. No custom Python required.
Prerequisitefamily: tabular in recipe.yaml · data.csv with a target column
method: linear | logistic | ridge | lasso | elasticnet
Linear models. Set lambda / l1_ratio as needed for ridge, lasso, elasticnet.
family: tabular
method: ridge
lambda: 1.0
data:
path: data.csv
target: y
eval:
metric: mse
method: tree | forest | boosting | gp
Trees, random forests, gradient boosting, and gaussian processes. Boosting library: auto | xgboost | lightgbm | catboost | sklearn.
family: tabular
method: boosting
library: auto
trees: 100
depth: 4
data:
path: data.csv
target: y
eval:
metric: accuracy