case classOneHotEncoder extends Product with Serializable
An immutable preprocessor that encodes categorical features as a one-hot (dummy) matrix made up of binary columns.
An immutable preprocessor that encodes categorical features as a one-hot (dummy) matrix made up of binary columns.
Preprocessor expects that categorical input values are in range [0, max(values)). If during the transformation
a value larger than during training is encountered it is ignored, i.e. no value is set in the binary encoded matrix.
Transformed categorical columns are appended at the end of the feature matrix.
case classStandardScaler extends Product with Serializable
An immutable preprocessor that transforms features by subtracting the mean and scaling to unit variance.
An immutable preprocessor that encodes categorical features as a one-hot (dummy) matrix made up of binary columns. Preprocessor expects that categorical input values are in range [0, max(values)). If during the transformation a value larger than during training is encountered it is ignored, i.e. no value is set in the binary encoded matrix. Transformed categorical columns are appended at the end of the feature matrix.