Predictive analysis with large predictive models

Abstract

An approach to optimizing predictive model analysis, comprising creating one or more model templates, decomposing a predictive model, wherein model information is extracted from the predictive model, storing the model information in the one or more model templates, creating a plurality of sub-models, associated with the predictive model, using the stored model information, sending the plurality of sub-models to a scoring engine, receiving results based on the plurality of sub-models from the scoring engine and generating predictions based on combining the results received from the scoring engine. The generated predictions can be sent to one or more analytic applications for further processing.

What is claimed is: 
     
         1 . A method for optimizing predictive model analysis, the method comprising:
 creating, by one or more computer processors, one or more model templates, wherein the one or more model templates are database tables stored in a model repository and comprise a master model template and an associated models template;   decomposing, by one or more computer processors, a predictive model, wherein model information is extracted from the predictive model;   storing, by one or more computer processors, the model information in the one or more model templates;   creating, by one or more computer processors, a plurality of sub-models, associated with the predictive model, using the stored model information, wherein the master model template comprises information which is common to the plurality of sub-models, and wherein the associated models template comprises a plurality of segments associated with the predictive model and different sub-models of the plurality of sub-models comprise different segments of the plurality of segments;   sending, by one or more computer processors, the plurality of sub-models to a scoring engine, wherein the scoring engine comprises one or more computing facilities and a subset of the plurality of sub-models is sent to a portion of the computing facilities based on a type of data the portion of the computing facilities are processing;   receiving, by one or more computer processors, results based on the plurality of sub-models from the scoring engine, wherein the results are combined based on at least one of a combination method associated with the predictive model and a default combination method;   generating, by one or more computer processors, predictions based on combining the results received from the scoring engine; and   sending, by one or more computer processors, the generated predictions to one or more analytic application for further processing.