Claim Missing Document
Check
Articles

Found 3 Documents
Search
Journal : Integra: Journal of Integrated Mathematics and Computer Science

Robust Panel Data Regression Analysis using the Least Trimmed Squares (LTS) Estimator on Poverty Line Data in Lampung Province Lestari, Windi; Widiarti; Utami, Bernadhita Herindri Samodera; Usman, Mustofa; Handayani, Vitri Aprilla
Integra: Journal of Integrated Mathematics and Computer Science Vol. 1 No. 2 (2024): July
Publisher : Magister Program of Mathematics, Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26554/integrajimcs.20241210

Abstract

Robust regression is an alternative method in regression analysis designed to produce stable parameter estimates, even when the data contain outliers or deviate from classical assumptions. One of its estimation techniques, the Least Trimmed Square (LTS),works by minimizing the smallest squared residuals, thereby assigning smaller weights to extreme data points. This method serves as a solution when classical approaches, such as Ordinary Least Squares (OLS), fail to meet the assumptions, especially in socio-economic data that are often complex and prone to outliers. This study employs robust regression with the LTS estimator on panel data to examine the impact of population size , population density , and registered job vacancies on poverty lines in Lampung Province. The data cover 15 districts and cities from 2019 to 2023. The analysis results show that the model obtained has a coefficient of determination of R2=0.8909. This means that the three predictor variables can explain 89.09% of the variation in the poverty line.
Georaphically Weighted Ridge Regression Modelling on 2023 Poverty Indicators Data in the Provinces of West Kalimantan and Central Kalimantan Anjani, Syarli Dita; Widiarti; Utami, Bernadhita Herindri Samodera; Usman, Mustofa; Handayani, Vitri Aprilla
Integra: Journal of Integrated Mathematics and Computer Science Vol. 1 No. 3 (2024): November
Publisher : Magister Program of Mathematics, Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26554/integrajimcs.20241320

Abstract

Regression analysis is a method to explain the relations between independent variables and a dependent variable. Linear regression analysis relies on certain assumptions, one of the assumption is homogeneity. However, there is a situation when the variance at each observation differs or called spatial heterogeneity.This issue can be solved using Geographically Weighted Regression (GWR), a statistical method that can be fixed spatial heterogeneity by adding a local weighted matrix, the result in GWR model is a local model for each observation point. However, GWR has a limitation, it cannot handle multicollinearity. Ridge regression is a method used to solved multicollinearity by adding a bias constant (λ). A GWR model that contains multicollinearity and fixed using ridge regression is known as Geographically Weighted Ridge Regression (GWRR).
Multidimensional Log-Linear Modeling (Case Study: Gender, Age, Head Circumference, and Nutritional Status Among Early Childhood Children) Yoka, Ranara Athalla; Usman, Mustofa; Chasanah, Siti Laelatul; Widiarti; Handayani, Vitri Aprilla
Integra: Journal of Integrated Mathematics and Computer Science Vol. 2 No. 2 (2025): July
Publisher : Magister Program of Mathematics, Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26554/integrajimcs.20252228

Abstract

Poor nutritional status tends to increase the risk of morbidity and mortality among children in developing countries. Therefore, data on these rates can be an important indicator in describing the condition of undernutrition in a community. Log-linear model analysis can be used to categorize data on nutritional status. Based on data obtained from the Rajabasa Indah Health Center area, Rajabasa Subdistrict, Bandar Lampung City, there are 418 children who have examined at the Posyandu. The analysis model conducted in this study involves four variables, each variable is categorized into several categories according to predetermined criteria. Gender with two categories (male and female), age with two categories (1-12 months and 13-60 months), head circumference with two categories (normal and abnormal), and nutritional status with three categories (undernourished, well-nourished, and overnourished). This study aims to determine the best model using log-linear analysis that can explain the relationship between the four variables. The results obtained are the best model for the data involved in the [UG][LG][J] structure, the structure describes the interaction between age and nutritional status and head circumference and nutritional status.