In this second of a series of three articles about the SAP Predictive Analytics toolset (previously marketed separately as SAP InfiniteInsight and SAP Predictive Analysis), learn about its Expert Analytics functionality. The first article covered the Automated Analytics functionality. This article focuses on how to use this more advanced toolset in the real world to discover hidden process improvements and potential profits.
Reading this article you will learn:
- About the target audience for SAP Predictive Analytics and its sub-components: Automated Analytics and Expert Analytics
- How to create an automated classification analytics tool using basic SAP Predictive Analytics’ Expert Analytics functions
SAP Predictive Analytics is a rapidly changing toolset used by data analysts to help them do their jobs more efficiently and accurately. With SAP Predictive Analytics, you can analyze such diverse topics as which bridge might collapse first or, as in the example used here, determine what type of customer might buy your products. SAP Predictive Analytics is really two unique experiences, with slightly different target audiences, rolled into one.
SAP Predictive Analytics is now one wrapper around what used to be two distinct tools, SAP InfiniteInsight and SAP Predictive Analysis. SAP Predictive Analytics incorporates SAP InfiniteInsight’s original functionality in its Automated Analytics tool, and the functions of Predictive Analysis are now included in the new Expert Analytics tool.
Automated Analytics Versus Predictive Analytics: Two Paths to a Common Goal
Automated Analytics is a toolset that SAP acquired from KXEN (previously marketed by both KXEN and SAP as InfiniteInsight). Expert Analytics is the SAP toolset that made up what was previously SAP Predictive Analysis.
In the first release, SAP Predictive Analytics 2.0, SAP created one shell that could launch either Automated Analytics or Expert Analytics. This shell allows companies to install just one tool—SAP Predictive Analytics 2.0—and deploy both experiences to use as they see fit. This is just a first step; tighter integration between these tools is planned for future releases, including the latest version, SAP Predictive Analytics 2.3. (Note: For the purposes of this article, I focus on the SAP Predictive Analytics 2.0 version.)
The main difference between these two tools is that the Automated Analytics tool’s graphical user interface (GUI) is more automated, but it offers fewer options. The Expert Analytics tool’s GUI, on the other hand, is less automated, but offers more power and flexibility. Other differences are in the actual algorithms used by each tool. When these tools were in separate products, the algorithms used were significantly different; this is less of an issue now. As of the 2.0 release some of the Automated Analytics’ algorithms can be deployed under the Expert Analytics mode.
What are some of the other differences between Automated Analytics and Expert Analytics? Automated Analytics is more what I call wizard based. It walks you through each screen in the process of data mining (i.e., getting the data, training, applying the model, and storing the results). You click Next and Next again to progress through the process.
The Expert Analytics GUI is less linear. It shows users the five phases in which to work in a typical data-mining project: Acquire Data, Predict Insights, Build Visualizations, Explore Data, and Share with your teammates. The subtasks that you can choose during each of these phases are more flexible than those offered by Automated Analytics.