Scoring with NMF

Non-Negative Matrix Factorization (NMF) can be used as a pre-processing step for dimensionality reduction in classification, regression, clustering, and other machine learning tasks. Scoring an NMF model produces data projections in the new feature space. The magnitude of a projection indicates how strongly a record maps to a feature.

The SQL scoring functions for feature extraction support NMF models. When the functions are invoked with the analytical syntax, the functions build and apply a transient NMF model. The feature extraction functions are: FEATURE_DETAILS, FEATURE_ID, FEATURE_SET, and FEATURE_VALUE.