Shreekant Shiralkar and Rohit Kumar Das show how to apply SAP Predictive Analysis Library (PAL) functions in SAP HANA to retail store data to calculate an optimum price.
Reading this article you will learn how to:
- Apply the Predictive Analysis Library (PAL) within SAP HANA for predicting the optimum price for a product by using k-means and multiple linear regression algorithms
- Create an SAP HANA calculation view, an SAP HANA procedure, and SAP HANA predictive modeling
The Predictive Analysis Library (aka PAL) is a set of predictive algorithmic functions embedded in SAP HANA. Each of the algorithms is written in native C++ and is capable of execution at the database layer (i.e., in-memory computing). Execution at the database layer enables computing of an extremely large set of data with millions of rows in a few seconds. You can generate results in real time.
The SAP HANA Predictive Analysis Library (PAL) contains the most popular and proven algorithms applicable to real-life scenarios. Each algorithm has a specific property that you can configure for its execution to meet the requirements of a particular business scenario.
We demonstrate how PAL enables better business decision making with a case study that shows how to apply regression analysis over data clusters for a pricing strategy in the retail industry. You can predict the right price of an article at a future time, providing for higher profitability. Application of the k-means cluster analysis and multiple linear regression (MLR) algorithm is explained through step-by-step instructions. You can create relevant clusters and find relationships between different sales parameters in an electronic retail store to arrive at the optimized pricing for a product.