In another move reflecting the shift to “mainstream analytics” – the predictive kind – SAS just announced the release of SAS Rapid Predictive Modeler. The product became generally available mid August.
The product allows less sophisticated users to create a basic model from within an Excel interface. SAS has added Rapid Predictive Modeler (RPM) as a task within the SAS Add-in for Microsoft Office. The SAS Office Add-In is one of the best on the market (see this blog and report) in its breadth of data access, but also, because it more often keeps the data on the server versus pulling everything into the spreadsheet. The Add-In also had a number of features that complement RPM, including the ability to create a sample data set, generate descriptive statistics, and to score a model. With RPM, a user can choose to create a basic, intermediate, or advanced model. A statistician can further tweak the model in any of the other SAS interfaces such as Enterprise Guide or Enterprise Miner. The first release of SAS Rapid Predictive Modeler supports only classification and regression models.
Of course the risk with RPM, as with any attempt to simplify analytics for non statisticians is that untrained users misapply the tool or misinterpret the results. However, anything that facilitates the dialogue between the business users (who know which variables are more likely to influence an outcome) with the statisticians (who know how to build the models) can only be a good thing.
Cindi Howson, BI Scorecard