Jurnal Aplikasi Statistika & Komputasi Statistik

Modeling the Stunting Prevalence Rate in Indonesia Using Multi-Predictor Truncated Spline Nonparametric Regression

2024-04-17T09:51:47+07:00

Introduction/Main Objectives: Stunting is the impaired growth and development that children experience from poor nutrition, repeated infection, and inadequate psychosocial stimulation. Background Problems: Based on data from the National Nutrition Status Survey (SSGI) in 2022, the prevalence of stunting in Indonesia was 21.6%, which is still above the WHO standard of below 20%. Novelty: This study was conducted with the aim of analysing the factors that influence the stunting prevalence rate in Indonesia using multi-predictor truncated spline nonparametric regression. Research Methods: The research data is secondary data taken from Health Statistics 2022 with response variables in the form of stunting prevalence. Finding Result: Based on the analysis, the best model to model the stunting prevalence rate is a multi-predictor truncated spline with three knots. In addition, it was found that four predictor variables which are the percentage of infants under 6 months old receiving exclusive breastfeeding, the average age of a mother's first pregnancy, the percentage of married women aged 15-49 using contraception, and the percentage of mothers who gave birth to a live child in the past two years and initiated early breastfeeding had a significant effect simultaneously and partially on the stunting prevalence rate in Indonesia.

Digital Literacy in Mediating the Influence of Education, Demography, Employment on Poverty

2024-03-26T14:30:56+07:00

This study investigated the influence of education, demography, and employment on poverty with digital literacy as a mediating variable. A structural Equation Modeling (SEM) with the Partial Least Square (PLS) method was applied. Significant indicators found are four indicators of education and digital literacy variables, two indicators of demographic and employment variables, and three indicators of poverty variables. It was found that education and employment variables had a significant influence on poverty with a negative influence. We found that no variable has a significant effect on digital literacy and there is no significant effect of digital literacy on Poverty.

Geographically Weighted Poisson Regression for Modeling the Number of Maternal Deaths in Papua Province

2024-03-28T11:43:53+07:00

Introduction/Main Objectives: Maternal Mortality Rate (MMR) in Indonesia is one of the main focuses in achieving the third Sustainable Development Goals (SDGs) in 2030. Background Problems: The Central Statistics Agency states that the MMR in Papua Province is the highest, reaching 565. Novelty: Given the diverse geographical conditions of each district/city in Papua Province, an analysis was carried out. Research Methods: Using the Geographically Weighted Poisson Regression (GWPR) method with the response variable being maternal mortality rates and variables predictors of health, social, and environmental factors. Finding/Results: Fixed Gaussian kernel GWPR is the best model with an AIC value of 27.6. Variable significantly influencing MMR include the percentage of households with access to adequate sanitation, the number of recipients of food assistance programs, and the number of doctors.

Application of Geographically Weighted Logistic Regression in Modeling the Human Development Index in East Java

2024-03-26T14:37:20+07:00

Main Objectives: pinpoint the factors influencing HDI, taking into consideration location and spatial factors. Problems: The Human Development Index (HDI) in East Java often fails to reflect actual conditions accurately, as disparities exist among districts and cities, with some falling below government expectations. Novelty: GWLR extends logistics regression by incorporating spatial factors, allowing for the identification of regional differences and influential factors affecting HDI based on actual data. Methods: To address this issue, the Geographically Weighted Logistic Regression (GWLR) method is employed. The independent variables used are Expected Years of Schooling (X1), Open Unemployment Rate (X2), and Morbidity Rate ( X3) in 2021, while dependent variable is the Human Development Index (Y). Finding/Results: The study reveals that GWLR provides a superior model compared to Ordinary Logistic Regression, indicated by a lower Akaike Information Criterion (AIC) of 28.72. Additionally, the GWLR model with Fixed Gaussian Kernel weights outperforms other weighting methods. At 90% confidence level, the significant variables influencing HDI are expected years of schooling (X1) and the open unemployment rate (X2). Given the relatively low HDI in Indonesia, the East Java Government should focus on improving these key areas to enhance HDI across districts and cities in the region.

Geographically Weighted Lasso Method in Modeling the Gross Regional Domestic Product of the Bali-Nusra Region

2024-04-23T08:46:44+07:00

Indonesia's Central Bureau of Statistics announced that economic growth in 2020 is still in the negative zone, and the group of provinces in the Bali-Nusra region has the most negligible impact on economic growth. The value of Gross Regional Domestic Product (GRDP) measures Indonesia's economic growth. GRDP is the total added value all regional business units generate at a particular time. This research aims to apply and interpret the results of the Geographically Weighted Lasso (GWL) method for GRDP in the Bali-Nusra region. The GWL method further develops the Geographically Weighted Regression (GWR) approach by adding the Least Absolute Shrinkage and Selection Operator (LASSO) method. The GWL method simultaneously selects insignificant variables by reducing the value of the regression coefficient to zero using the LASSO method. The data used has the effect of spatial heterogeneity and multicollinearity, a prerequisite for modeling with the GWL method. Based on the analysis conducted, there are 41 different GRDP models for each district/city in the Bali-Nusra region. The resulting GWL model provides a coefficient of determination of 95.84 % so that the resulting model can be used and is considered valid.

Modeling Multi-Output Back-Propagation DNN for Forecasting Indonesian Export-Import

2024-04-02T12:16:20+07:00

Introduction/Main Objectives: International trade through the mechanisms of exports and imports plays a significant role in the Indonesian economy, making the timely availability of export and import value data crucial. Background Problems: Export and import values are influenced by inflation and exchange rate factors. Novelty: This study identifies two categories of variables, namely output (export value and import value) and input (inflation rate and the exchange rate of the Rupiah against the US Dollar). Research Methods: the research approach utilizes a Multi-output Deep Neural Network (DNN) with a Back-propagation algorithm to model the input-output relationship. The method can provide forecasting results for two or more bivariate or multivariate output variables. Finding/Results: The modeling analysis results indicate that the optimal model network structure is DNN (3.4). This model successfully predicts output 1 (export value) and output 2 (import value) with Mean Absolute Percentage Error (MAPE) rates of 13.76% and 13.63%, respectively. Additionally, the forecasting results show predicted export and import values for November to be US$ 16,208.13 billion and US$ 15,105.33 billion, respectively. These findings offer important insights into the direction of Indonesia's international trade movement, which can serve as a basis for future economic decision-making.

Performance Study of Prediction Intervals with Random Forest for Poverty Data Analysis

2023-12-20T15:13:01+07:00

Introduction/Main Objectives: Determine the prediction interval with for analyzing poverty data at the Regency/City level in Indonesia. Background Problems: Poverty will be a topic in various discussion and debates in the future. Novelty: This study’s methods for constructed prediction intervals are LM, Quant, SPI, HDR, and CHDR. This method can improve the prediction interval performance with Random Forests. Research Methods: The method for building forests and obtaining BOP in this study is CART with the LS splitting rule. Finding/Results: The results of this study are that the best method for one replication is HDR with 500 trees. The best method for 100 repetitions is LM. Based on hypothesis testing, there is sufficient evidence to say no difference between the LM, SPI, Quant, HDR, and CHDR methods for 100 replications at a 5% significance level.