Journal of Data Analysis

  • Home
  • About
  • Login
  • Register
  • Categories
  • Search
  • Current
  • Archives
  • Announcements
  • PUBLICATION ETHICS
Home > Volume 3, Number 1, June 2020 > Junaidi
User
 
Template of Article
Example of Article
Certificate of Reviewing
Open Journal Systems
PRINCIPAL CONTACT
Asep Rusyana
Department of Statistics, Faculty of Mathematics and Natural Sciences, Syiah Kuala University
Jalan Syech Abdurrauf No.3, Kopelma Darussalam, Banda Aceh 23111, Aceh, Indonesia
Email: jda@unsyiah.ac.id
Mobile Phone: +6281360635965

Picture Link:

 

Link
  • Department of Statistics
  • Forstat
  • Universitas Syiah Kuala
  • ISSN Online
  • Google Scholar
  • Garuda
  • Crossref
  • IOS
  • WorldCat
  • BASE
  • ISJD
  • Microsoft Academic
  • DOAJ
  • Arjuna Akreditasi Jurnal Nasional
  • Sinta
  • Scopus
  • Web of Science

 Picture Link:

 

 

 

Journal Help

Analisis Sensitivitas Model Regresi Linier Berganda Menggunakan Pendekatan Bayesian (Distribusi Prior Normal)

Junaidi Junaidi, Mohammad Fajri, Yandi Ristawan

Abstract

Metode regresi linier berganda merupakan metode yang memodelkan hubungan antara peubah respon (y) dan beberapa peubah predictor (x). Pada metode Bayesian parameter yang digunakan merupakan variabel random yang dilkukan dengan mengalikan Likelihood dengan distribusi prior. Distribusi prior adalah distribusi subyektif berdasarkan pada keyakinan seseorang dan dirumuskan sebelum data sampel diambil. Tujuan penelitian ini adalah  untuk menganalisis sensitivitas dari parameter-paremeter pada model regresi linier berganda yang akan dilakukan dengan menggunakan prior berdistribusi Normal. Selanjutnya, penerapan model pada data aset bank di Indonesia dengan hasil estimasi parameter yaitu , , , , , dan , dengan selang kepercayaan 95%  untuk setiap parameter yang dihasilkan yaitu==       (-1,427 ; 3,594),  =(-5,07;0,3061), =(, , dan  = (-0,5955 ; 2,487). Nilai estimasi parameter yang diperoleh dengan pendekatan Bayesian mendekati nilai parameter yang diperoleh dengan Frequantis. Selang kepercayaan yang diperoleh juga mendekati dengan hasil frequentis yang memiliki interval lebih sempit dibandingkan nilai interval dengan metode OLS. Hal ini menunjukkan bahwa metode Bayesian merupakan suatu pendekatan yang dapat digunakan untuk mengestimasi parameter pada analisis regresi linier berganda.

 

The multiple linear regression method is a method that models the relationship between the response variable (y) and several predictor variables (x). In the Bayesian method, the parameters used are random variables which are conducted by multiplying the likelihood with the prior distribution. The prior distribution is a subjective distribution based on a person's beliefs and is formulated before the sample data is taken. The purpose of this study is to analyze the sensitivity of the parameters in the multiple linear regression model that will be carried out using prior normal distribution. Furthermore, the application of the model to the data on bank assets in Indonesia with the results of parameter estimation is β0 = 23.06, β1 = 1.05, β2 = -2,379, β3 = -0,4786, β4 = -0.03796, and β5 = 0.9075, with a 95% confidence interval for each resulting parameter, namely β0 = (6,052; 40,200), β1 = (-1,427; 3,594), β2 = (- 5.07; 0, 3061), β3 = (0.9896; 0.03289), β4 = (- 1,224; 1.139), and β5 = (-0.5955; 2.487). The parameter estimate value obtained by the Bayesian approach is close to the parameter value obtained by Frequantis. The confidence interval obtained is also close to the frequentis result which has a narrower interval than the interval value with the OLS method. This shows that the Bayesian method is an approach that can be used to estimate parameters in multiple linear regression analysis.

 Keywords

Analisis Regresi; Bayesian; Sensitivitas; Aset Bank

 Full Text:

PDF (Bahasa Indonesia)

References

Kesuma, Z. M., Rusyana A., dan Rahayu L. 2020. Factors affecting adolescent nutritional status in Banda Aceh, Indonesia. Journal of Physics: Conference Series. 1490, 012049. doi:10.1088/1742-6596/1490/1/012049.

Sulistyono dan Sulistyowati, W. 2017. Peramalan produksi dengan regresi linier berganda. Prozima ,vol.2. Sidoarjo. Universitas Sidoarjo.

Marzuki, Sofyan, H. dan, Rusyana, A. 2010. Pendugaan Selang Kepercayaan Persentil Bootstrap Nonparametrik untuk Parameter Regresi. Statistika. Vol. 10, No.1, 13-23. Universitas Islam Bandung, Bandung.

Bain, L. J. dan Engelhardt, M. 1992. Introduction to Probability and Mathematical Statistics. Second Edition. California. Duxbury Press.

Junaidi, Nur, D. dan Stojanovski, E. 2011. Prior Sensitivity Analysis for a Hierarchical Model. In Proceedings of the Fourth Annual ASEARC Conference, Paper 13, University of Western Sydney, Paramatta, NSW, Australia, 17–18 February 2011. http://ro.uow.edu.au/asearc/24.

Utami, T., Wahyu, A. R. dan Prahutama, A. 2016. Pemodelan Regresi Berganda Dan Geographically Weighted Regression Pada Tingka Pengangguran Terbuka Di Jawa Tengah. Media Statistika. Universitas Diponegoro.

Evans, S. 2012. Bayesian Regression Analysis. Faculty of The College of Arts and Sciences, University of Louisville.

Walpole, R. E dan Myers, R. H. 1986. Ilmu Peluang dan Statistika untuk Insinyur dan Ilmuwan, Terbitan Kedua. Bandung. ITB.

Horst, R. 2009. The Weibull Distribution A Handbook. Jerman. Justua-Liebig-University Giessen.

Gelman, A., Carlin, J. B., Sternhal, S. and Rubin, D. B. 2004. Bayesian Data Analysis. Second edition. New York. Chapman & Hall.

Box, G. E. P. dan Tiao, G. C. 1973. Bayesian Inference In Statistical Analysis. Philippines. Addision-Wesley Publishing Company.

Bolstad, W. M. 2007. Introduction to Bayesian Statistics. Second Edition. New Jersey. John Wiley & Sons Inc.

Junaidi, Magdalena, M. dan Nurdiono 2013. Transparansi Informasi Suku Bunga Dasar Kredit Pada Kredit UMKM. Indonesia Accounting Research Journal, Vol. 1, No. 2. Yogyakarta.

Ainul, A. M., Junaidi dan Utami, I. T. 2018. Penerapan Analisis Regresi Linier Berganda dengan pendekatan Bayesian pada data Aset Bank di Indonesia. JUTEKS. LPPM UNHAS.

DOI: https://doi.org/10.24815/jda.v3i1.18358

Refbacks

  • There are currently no refbacks.

Partnership with: 

About The Authors

Junaidi Junaidi
Department of Statistics, Tadulako University, Palu, Indonesia
Indonesia

Mohammad Fajri
Department of Statistics, Tadulako University, Palu, Indonesia.
Indonesia

Yandi Ristawan
Department of Statistics, Tadulako University, Palu, Indonesia.
Indonesia

Indexed by:

 
Keywords ARIMA Analisis Regresi Banda Aceh Biplot Canonical correlation analysis Computer Network Correspondence Development areas Forecasting Hybrid Kemiskinan Korelasi MANOVA Mean lingkage Multidimensional Nutritonal status Quality of Service Software-defined Network Spatial Regression Stunting Sumatera Island
Journal Content

Browse
  • By Issue
  • By Author
  • By Title
  • Other Journals
  • Categories
Since 17 March 2018
Visitors since 17 March 2018

Flag Counter

View My Stats

Font Size

 

 

 ISSN 2623-0658 (Print) 

 ISSN 2623-2286 (Online)

 

Creative Commons License
Journal of Data Analysis (JDA) is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.