A New Approach to Robust Partial Least Squares Regression Analysis
Journal Title: INTERNATIONAL JOURNAL OF MATHEMATICS TRENDS AND TECHNOLOGY - Year 2014, Vol 9, Issue 3
Abstract
Partial Least Squares Regression (PLSR) is a linear regression technique developed to relate many independent variables to one or several dependent variables. Robust methods are introduced to reduce or remove the effects of outlying data points. In the previous studies in robust PLSR field it has been mentioned that if the sample covariance matrix is properly robustified further robustification of the linear regression steps of the PLS1 algorithm (PLSR with univariate dependent variable) becomes unnecessary. Therefore, the purpose of this study is to propose a new approach to robust PLSR based on statistical procedures for covariance matrix robustification by selecting the well-known S-estimators. Both simulation results and an analysis on a real data set, which is used in robust PLSR literature frequently, showing the effectiveness, success in fitting to regular data points and predictive power of the new proposed robust PLSR method.
Authors and Affiliations
Esra Polat , Suleyman Gunay
Keywords
Related Articles
- EP ID EP152632
- DOI -
- Views 252
- Downloads 0
Pattern Generation for Two Dimensional Cutting Stock Problem
Selection of feasible cutting patterns in order to minimize the raw material wastage which is known as cutting stock problem has become a key factor of the success in today’s competitive manufacturing industries. I...
MHD Effects on Composite Slider Bearing Lubricated with Couple-stress Fluids
A magneto-hydrodynamic wide composite slider bearing lubricated with couple-stress fluids is numerically analysed. A modified Reynolds equation has been derived to account for the transverse magnetic field in the couple-...
Multicast Multi-Path Power Efficient Routing In Mobile Adhoc Networks
The proposed work is aimed to provide efficient power aware routing considering real and non real time data transfer. Ad hoc networks are peer to peer, autonomous networks comprised of wireless mobile devices. Th...
Higher Six Dimensional Plane Gravitational Waves in Bimetric Relativity
In this paper, I studied the, type plane gravitation waves for higher six dimensions and it will observed that the result for vacuum space and formatter cosmic strings respectively.
On Edge Pair Sum Labeling of Graphs
An injective map f : E(G) → {±1,±2,…,±q} is said to be an edge pair sum labeling of a graph G(p, q) if the induced vertex function f^*:V(G)→Z-{0} defined by f^* (v)=∑_(e∊E_v)▒f(e) is one – one, where E_v denotes the set...