Model

Base models

Abstractions for training models with AutoML.

AutoML

Establish a champion model for a dataset with AutoML

model = AutoMLModel()
model.fit(df, target='col_name')

AutoML

AutoTS

Establish a time-series champion model across multiple forecast horizons with AutoTS

model = AutoTSModel()
model.fit(df, target='col_name', datetime_partition_column='date_col')

AutoTS

AutoAnomaly

Train diverse anomaly detection models with AutoML

model = AutoAnomalyModel()
model.fit(df)

AutoAnomaly

AutoCluster

Train diverse clustering models with AutoML

model = AutoClusterModel(n_cluster=[2, 3, 5])
model.fit(df)

AutoClustering

Model wrappers

Abstractions for working with base models within a single DR project.

FeatureDiscovery

Model with and discover features across multiple datasets PROTOTYPE

base_model = drx.AutoMLModel()
model = drx.FeatureDiscoveryModel(base_model)
rltn = drx.Relationship(df2, keys=[key1, key2])
model.fit(df1, target="target", relationship_configuration=rltn, feature_engineering_prediction_point="Date")

FeatureDiscovery

ColumnReduce

Fit models on iteratively column-reduced training data PROTOTYPE

base_model = AutoMLModel()
model = ColumnReduceModel(base_model)
model.fit(df, target='col_name')

ColumnReduce

SelfDiscovery

Perform feature discovery on a single dataset using a self join PROTOTYPE

base_model = AutoMLModel()
model = SelfDiscoveryModel(base_model)
model.fit(df, target='col_name', keys='my_join_key')

SelfDiscovery