Estimation of spline function in nonparametric path analysis based on penalized weighted least square (PWLS)

Open

Adji Achmad Rinaldo Fernandes, Solimun, Endang Arisoesilaningsih

2017 AIP Conference Proceedings Vol. 1913 Conference paper Cited by 11 Quartile

Abstract

The aim of this research is to estimate the spline in Path Analysis-based on Nonparametric Regression using Penalized Weighted Least Square (PWLS) approach. Approach used is Reproducing Kernel Hilbert Space at sobolev space. Nonparametric path analysis model on the equation y1i=f1.1(x1i)+ϵ1i;y2i=f1.2(x1i)+f2.2(y1i)+ϵ2i; i=1,2,..,n Nonparametric Path Analysis which meet the criteria of minimizing PWLS minfw.kW2m[aw.k,bw.k],k=1,2{ (2n)-1(y-f)TΣ-1(y-f)+k=12w=12λw.kaw.kbw.k [fw.k(m)(xi)] 2dxi } is f=Ay with A=T1(T1TU1-1-1T1)-1T1TU1-1-1+V1U1-1-1[I-T1(T1TU1-1-1T1)-1T1TU1-1-1]+T2(T2TU2-1-1T2)-1T2TU2-1-1+V2U2-1-1[I1-T2(T2TU2-1-1T2)-1T2TU2-1-1] © 2017 Author(s).

Affiliations

Statistics Study Program, Department of Mathematics, Faculty of Mathematics and Natural Sciences, Brawijaya University, Indonesia; Department of Biology, Faculty of Mathematics and Natural Sciences, Brawijaya University, Indonesia