This IBMAr RedpaperTM publication presents the process and steps that were taken to move from an R language forecasting solution to an IBM SPSSAr Modeler solution. The paper identifies the key challenges that the team faced and the lessons they learned. It describes the journey from analysis through design to key actions that were taken during development to make the conversion successful. The solution approach is described in detail so that you can learn how the team broke the original R solution architecture into logical components in order to plan for the conversion project. You see key aspects of the conversion from R to IBM SPSS Modeler and how basic parts, such as data preparation, verification, pre-screening, and automating data quality checks, are accomplished. The paper consists of three chapters: Chapter 1 introduces the business background and the problem domain. Chapter 2 explains critical technical challenges that the team confronted and solved. Chapter 3 focuses on lessons that were learned during this process and ideas that might apply to your conversion project. This paper applies to various audiences: Decision makers and IT Architects who focus on the architecture, roadmap, software platform, and total cost of ownership. Solution development team members who are involved in creating statistical/analytics-based solutions and who are familiar with R and IBM SPSS Modeler.It is easy to conduct an ANOVA test by calling anova() in R. And, the results that are returned by the anova() function contain a p-value. However, implementing the same algorithm in SPSS Modeler is difficult. First, no separate node isanbsp;...
|Title||:||Our Experience Converting an IBM Forecasting Solution from R to IBM SPSS Modeler|
|Author||:||Pitipong JS Lin, Fan Li, Yin Long, Stefa Etchegaray Garcia, Jyotishko Biswas, IBM Redbooks|
|Publisher||:||IBM Redbooks - 2015-03-06|