SPARSEML Procedure

The SPARSEML procedure supports the use of sparse data as input data. It implements the support vector machines (SVM) algorithm in SAS Viya for binary classification. By using the coordinate descent optimization method, the SPARSEML procedure can train large-scale data sets that can be both wide and deep.

Like other predictive modeling tools in SAS Viya, the SPARSEML procedure uses input data to train a model and generates an analytic store that can be deployed through the ASTORE procedure. It can load data from multiple nodes and perform computations in parallel.