PERBANDINGAN METODE REGRESI LINIER MULTIVARIABEL DAN REGRESI SPLINE MULTIVARIABEL DALAM PEMODELAN INDEKS HARGA SAHAM GABUNGAN

Ihdayani Banun Afa -  , Indonesia
*Suparti Suparti -  Departemen Statistika, Universitas Diponegoro, Indonesia
Rita Rahmawati -  Departemen Statistika, Universitas Diponegoro, Indonesia
Received: 21 Aug 2018; Published: 30 Dec 2018.
Open Access Copyright (c) 2018 MEDIA STATISTIKA
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Citation Format:
Article Info
Section: Articles
Language: ID
Full Text:
Statistics: 243 545
Abstract

The composite stock price index or Indonesia Composite Index (ICI) is a composite index of all stocks listed on the Indonesia Stock Exchange and its movements indicate conditions that occur in the capital market. For investors, the ICI movement is one of the important indicator to make a decision whether the stocks will be sold, held or bought new shares. The ICI movement (y) was influenced by several factors including Inflation (x1), Exchange Rate (x2) and SBI interest rate (x3). This study aims to compare the ICI modeling  using the parameric and nonparametric approaches, namely multivariable linear regression and multivariable spline regression. Determination of the better model is based on the smaller MSE and the larger R2. The best regression model is multivariable spline regression with x1, x2 and x3, each with a sequence orde (3,2,2) and the number of knot points (1,2,2).

Keywords: Indonesia Composite Index, Multiple Linear Regression, Multivariable Spline Regression, MSE, R2

Article Metrics: