The MARS procedure builds flexible regression models by fitting separate splines (or basis functions) to distinct intervals of the predictor variables. Both the variables to use and the end points of the intervals for each variable-referred to as knots-are found via a brute force, exhaustive search procedure, using very fast update algorithms and efficient program coding. Variables, knots and interactions are optimized simultaneously by evaluating a "loss of fit" (LOF) criterion. MARS chooses the LOF that most improves the model at each step. In addition to searching variables one by one, MARS also searches for interactions between variables, allowing any degree of interaction to be considered.
Did you find this white paper useful?
25 out of 46 users found this white paper useful
Related white papers
End-to-End Application Performance Management
The particular challenge of maintaining complex applications is that they cut across separate IT silos. An application will often include a Web tier, application logic, a database, and a storage...
Software Estimation Best Practices – Level 200  Â
With a software project, the top three factors that determine its ultimate success are estimation, estimation, and estimation. Without good estimation, a good plan is impossible because your plan is...
Design and Architecture for .NET Applications - Level 300
There are many ways to design a .NET application. In this webcast based on MagenicÂ’s architecture experience with .NET at numerous clients, learn which ways work best for Web and...
Software Developer Increases Product Scalability With Access to the Right Resources
Multi-Systems, Inc. (MSI) wanted to improve the scalability of its next-generation hotel property management system. The company attended an Architecture Design Session and Proof-of-Concept Workshop at a Microsoft Technology Center,...
Application Lifecycle Management With ClearQuest 7.1.0.0
This overview of the concepts and design goals behind an out-of-the-box Application Lifecycle Management (ALM) solution for IBM Rational ClearQuest illustrates the benefits of using ClearQuest and the ALM package...
MSDN Webcast: Overview of New Features for Windows Applications (Level 200)
This webcast is specifically for a Windows application developer. The presenter of this webcast takes a fresh look at how the functionality in the Microsoft Visual Studio 2008 development system...
Top Strategies to Reduce Cost While Increasing Quality in Application Delivery
Reducing cost and increasing quality of applications at the same time is not a combination you see very often. IT usually has to sacrifice one for the other. In this...
