7.1.14 Model Types

AutoML supports the following training models. When training AutoML a model, use the ML_TRAIN model_list and exclude_model_list options to specify the training models to consider or exclude. The Model Metadata includes the algorithm_name field, which defines the model type.

Classification Models

• LogisticRegression

• GaussianNB

• DecisionTreeClassifier

• RandomForestClassifier

• XGBClassifier

• LGBMClassifier

• SVC

• LinearSVC

• ExtraTreesClassifier

Regression Models

• DecisionTreeRegressor

• RandomForestRegressor

• LinearRegression

• LGBMRegressor

• XGBRegressor

• SVR

• LinearSVR

• ExtraTreesRegressor

Forecasting Models

Univariate endogenous models:

• NaiveForecaster

• ThetaForecaster

• ExpSmoothForecaster

• ETSForecaster

• STLwESForecaster: STLForecast with ExponentialSmoothing substructure

• STLwARIMAForecaster: STLForecast with ARIMA substructure

Univariate endogenous with exogenous models:

• SARIMAXForecaster

• OrbitForecaster

Multivariate endogenous with exogenous models:

• VARMAXForecaster

Univariate or multivariate endogenous with exogenous models:

• DynFactorForecaster

Anomaly Detection Models

• GkNN: Generalized kth Nearest Neighbors

• PCA: Principal Component Analysis

• GLOF: Generalized Local Outlier Factor

Recommendation Models

The TwoTower model is the default model and can generate recommendations with implicit or explicit feedback. See Recommendation Training Models to learn more.

Recommendation models that rate users or items to use with explicit feedback:

• Baseline

• CoClustering

• NormalPredictor

• SlopeOne

• Matrix factorization models:

  • SVD

  • SVDpp

  • NMF

Recommendation models that rank users or items to use with implicit feedback:

• BPR: Bayesian Personalized Ranking from Implicit Feedback

• CTR: Collaborative Topic Regression