UNP Journal of Statistics and Data Science

K-Means Clustering of Jambi Province Based on Economic Growth in 2023

Fathina Nafisa Putri, Dina Fitria, Admi Salma — Sat, 28 Feb 2026 00:00:00 +0000

Economic growth describes a region’s economic condition. In Jambi Province, although recovery after the COVID-19 pandemic has been visible, gaps between districts and cities still exist due to income inequality, poverty, unemployment, and differences in human capital quality shown by the Human Development Index. This study aims to group districts/cities in Jambi Province based on economic growth and its determinants using the k-means clustering method. The analysis resulted in five clusters with distinct characteristics. Cluster 1, located in the central region, is characterized by relatively low economic growth and human capital, along with a high poverty rate. Cluster 2, covering areas in the western highlands and eastern region, shows strong human capital and a low poverty rate. Cluster 3, in the western part of the province, is marked by low poverty and unemployment rates. Cluster 4, situated in the northeastern coastal area, has the highest Gross Regional Domestic Product (GRDP) per capita and the lowest unemployment rate but struggles with a high poverty rate and weak human capital. Meanwhile, Cluster 5, representing the provincial capital area, demonstrates robust economic growth and strong human capital, although unemployment remains a key issue. These findings highlight the heterogeneity of regional conditions, suggesting that development policies must be tailored to each cluster to promote inclusive growth and equitable welfare.

Forecasting Smallholder Oil Palm Yield in Riau Province through the SARIMA Approach

Septrina Kiki Arisandi, Dony Permana, Tessy Octavia Mukhti — Sat, 28 Feb 2026 00:00:00 +0000

Oil palm stands as one of Indonesia’s major agricultural sectors that plays a vital role in regional economic growth, particularly within Riau Province. However, its production often fluctuates due to seasonal and environmental factors, making accurate forecasting essential for planning and policy formulation. This study aims to forecast smallholder oil palm production in Riau Province through the Seasonal Autoregressive Integrated Moving Average (SARIMA) Approach. The data consist of monthly oil palm production from January 2006 to December 2023 obtained from the Central Bureau of Statistics (BPS) of Riau Province. The modeling process includes identifying the model structure, estimating parameters, performing diagnostic checks, and evaluating forecasting accuracy using the Mean Absolute Percentage Error (MAPE). The best model selected was SARIMA (2,0,0)(0,1,1)[12] with an AIC value of 4980.12 and a MAPE of 11.27%, indicating a good level of accuracy. The model effectively captured both seasonal and long-term trend patterns in production. The forecast results suggest that peak production typically occurs in August–September, while the lowest occurs in February–March. The study concludes that the SARIMA model provides a robust statistical framework for predicting oil palm production and can be applied as a decision-support tool in agricultural and economic planning for the province

Forecasting the Consumer Price Index of Padang City in 2024 using the Autoregressive Integrated Moving Average Method.pdf

Sat, 28 Feb 2026 00:00:00 +0000

The Consumer Price Index (CPI), which changes, is influenced by fluctuations in the prices of goods and services in Padang City every year. This is triggered by various factors that are of primary concern to the government. This study uses the Autoregressive Integrated Moving Average (ARIMA) forecasting method to forecast CPI in 2024 by relying on monthly data on the Padang City CPI for the period 2020 to 2023 obtained from BPS. This analysis identifies the ARIMA model (0,2,1) as the best and most optimal model based on the AIC and BIC values, does not show any autocorrelation, and is normally distributed. The forecasting model used shows a smooth and stable increase in the CPI in the period from January to December 2024. This model provides a positive signal for people's purchasing power and economic stability in Padang City in 2024. The results obtained are expected to be used as a strategic tool for preparing future goods and services price planning with more precision.

Classification of Tuberculosis in Rumah Sakit Paru Sumatera Barat Using the C5.0 Algorithm

Meliani Maya Sari, Zilrahmi, Dony Permana, Dwi Sulistiowati — Sat, 28 Feb 2026 00:00:00 +0000

Tuberculosis (TB) remains a serious public health problem, including in West Sumatra Province, where the number of reported cases has continued to increase in recent years. Consequently, effective methods are required to support early detection and accurate classification of TB patients. This study aims to classify the tuberculosis status of patients at Rumah Sakit Paru Sumatera Barat by applying the C5.0 algorithm. The data used in this study consists of secondary data extracted from patient medical records collected from october to december 2024 with a total of 150 patient medical records. The dataset included eight predictor variables representing clinical symptoms and one target variable, namely sputum smear (BTA) examination results. The research process involved data preprocessing, after which the dataset was divided into training and testing subsets using a 70:30 ratio, a classification model was developed using the C5.0 algorithm, and its performance was evaluated using a confusion matrix. The findings indicate that the C5.0 algorithm achieved an accuracy of 91.11%, with a precision of 95.83%, sensitivity of 88.46%, and specificity of 94.74%. Night sweats were identified as the most influential variable in the construction of the decision tree. These findings indicate that the C5.0 algorithm demonstrates excellent performance and can be applied as a decision support method for classifying tuberculosis based on patients’ clinical symptoms

Forecasting the Export Value of West Sumatra Province Using the Autoregressive Integrated Moving Average Method

Faddiah Gusti Handayani, Fadhilah Fitri, Dina Fitria — Sat, 28 Feb 2026 00:00:00 +0000

The export sector in Indonesia is a key driver of national economic growth, particularly through increased foreign exchange earnings and regional development. West Sumatra is one of the provinces that notably contributes to the country's export performance due to its abundant natural resources. This research aims to forecast export values for the upcoming 16 months, spanning from September 2025 to December 2026. The study employs the ARIMA method, which is suitable for various time-series patterns, including those involving non-stationary data. Based on the analysis, the ARIMA (3,1,0) model is identified as the most suitable, achieving a MAPE of 3.90%. The forecast indicates a slight downturn from August to September 2025, followed by a steady upward trend through December 2026, reflecting a stable and positive export outlook. The findings of this research are expected to provide valuable insights for local governments and industry stakeholders in designing more effective export policies.

Comparison of K-Means and Ward Methods in Clustering Indonesian Provinces Based on Household Basic Service Access

Sat, 28 Feb 2026 00:00:00 +0000

Disparities in household basic service access across provinces in Indonesia remain a key issue in regional development. Basic services such as access to improved drinking water, proper sanitation, electricity, and adequate housing are essential indicators of household welfare, making regional classification necessary to identify similarities and disparities among provinces. This study aims to cluster Indonesian provinces based on household basic service access indicators and to compare the performance of the K-Means method and Hierarchical Clustering using the Ward approach. The analysis was conducted using numerical data with Euclidean distance as a measure of similarity. The optimal number of clusters was determined using the Silhouette plot and further validated using the Silhouette Coefficient. The results indicate that both K-Means and Ward methods produce two optimal clusters representing provinces with relatively high and relatively low levels of household basic service access. Centroid analysis reveals clear differences between clusters across all indicators, particularly in electricity access and sanitation. Furthermore, the evaluation of clustering quality shows that the Ward method yields a higher Silhouette Coefficient than the K-Means method, indicating more compact clusters and better separation between clusters. Therefore, the Ward method is considered more effective in mapping patterns of household basic service access across provinces. The findings of this study can support regional planning by providing a clearer understanding of disparities in household basic service access in Indonesia.

Stress Analysis of FMIPA UII Students in Practicum Report using Perceived Stress Scale and Robust Regression

Abdullah Kafabih -, Fahma Zuaf Zarir, Naufal Fahrezi -, Edy Widodo — Sat, 28 Feb 2026 00:00:00 +0000

Higher education requires students not only to master theoretical knowledge but also to apply concepts through practicum activities. At FMIPA UII, the preparation of practicum reports often becomes a source of pressure due to the large number of reports, tight deadlines, and the complexity of data analysis. This study aims to measure students’ stress levels during practicum report preparation using the PSS-10 and to analyze the effects of the number of reports, semester level, and organizational involvement. Primary data were collected from students of all study programs at FMIPA UII through a questionnaire survey. The analysis was conducted using Ordinary Least Squares (OLS) for assumption testing and subsequently robust regression (Huber M-estimation) due to the presence of heteroskedasticity and influential outliers. Descriptive results indicate an average PSS score of 17.95, categorized as moderate stress. However, the robust regression results show that the number of reports, semester level, and organizational involvement do not have a significant effect, either simultaneously or partially, on academic stress. These findings suggest that student stress is more likely influenced by other factors such as time management, coping strategies, social support, practicum design, and overall academic workload.

Analysis of the Open Unemployment Rate on Poverty in Java in 2024 Using Smoothing Spline Regression

Nur Leli, Fadhilah Fitri, Nonong Amalita — Sat, 28 Feb 2026 00:00:00 +0000

Poverty and unemployment are two major issues in economic development that are interrelated and remain a serious concern in Indonesia. Java Island, as the center of economic activity and population in Indonesia, contributes relatively significantly to the national economy, but still faces issues of welfare inequality, including high unemployment rates in several regions and the persistence of people living below the poverty line. Therefore, analyzing the relationship between the Open Unemployment Rate and the Percentage of the Poor in Java Island is important to understand the socio-economic dynamics that occur. The analysis was carried out using the nonparametric regression method with a smoothing spline estimator. Based on the analysis results, an optimum model was obtained with a value of lambda of 0.04829734. The smoothing spline curve shows a negative relationship pattern, where an increase in the Open Unemployment Rate is followed by a decrease in the percentage of the poor. The Mean Square Error (MSE) value of 11.31277 indicates that the model has a relatively moderate level of prediction error and is able to represent the relationship pattern between variables quite well.

Cluster Analysis of Earthquakes on the Island of Sumatera in 2024 Using the DBSCAN Method

Zahrani Asyati Zulika, Yenni Kurniawati — Sat, 28 Feb 2026 00:00:00 +0000

Earthquakes are one of the most destructive and unpredictable natural disasters. Sumatera Island, being located along the Semangko Fault, typically experiences seismic movement due to contact between the Indo-Australian plate and the Eurasian plate. In this study, the DBSCAN method classifies earthquake incidents in Sumatera in 2024 into magnitude and depth categories. The data set, collected by the Meteorology, Climatology, and Geophysics Agency (BMKG), includes 163 earthquake events that occurred in Sumatera Island during 2024. The clustering process identified two main clusters: one representing deep earthquakes in inland areas and another consisting of shallow earthquakes along the western offshore region, near the megathrust zone. The Silhouette Coefficient was used to verify the clustering outcome, and the result was 0,58, which verifies a good formation of clusters. These findings provide insights into seismic patterns in Sumatera and can support disaster mitigation efforts.

Monthly Rainfall Forecasting in Pesisir Selatan Regency Using the Autoregressive Integrated Moving Average (ARIMA) Model

Nisa Ulhusna, Sulistiowati Dwi, Fitri Fadhilah — Tue, 03 Mar 2026 00:00:00 +0000

Rainfall is a climate variable that plays a crucial role in agricultural planning, water resource management, and hydrometeorological disaster mitigation. Therefore, a forecasting method capable of adequately describing the temporal patterns of rainfall data is required. This study aims to forecast monthly rainfall in Pesisir Selatan Regency using the Autoregressive Integrated Moving Average (ARIMA) method. The data used in this study are monthly rainfall data for the period 2015–2024. The analysis stages include missing data imputation, Box–Cox transformation, stationarity testing using the Augmented Dickey–Fuller (ADF) test, model identification through ACF and PACF plots, parameter estimation, and model evaluation based on the Akaike Information Criterion (AIC), residual diagnostic tests, and forecasting accuracy using Mean Absolute Percentage Error (MAPE). The results show that the ARIMA(0,1,1) model is the best model, as indicated by the lowest AIC value and residuals that satisfy the white noise assumption. The forecasting accuracy evaluation yields a MAPE value of 55.05%, indicating that the model’s ability to capture monthly rainfall variability is still limited. Rainfall forecasting for the period January to December 2025 produces relatively constant forecast values, reflecting the limitations of the ARIMA(0,1,1) model in representing seasonal variations. Therefore, this model is more suitable as a baseline approach for rainfall forecasting in Pesisir Selatan Regency. Future studies are recommended to apply models that incorporate seasonal components or external variables to improve forecasting accuracy.

Analysis of The Influence of Job Resources and Leadership Quality on Job Satisfaction Using Structural Equation Modeling

Azizah Apriyerni, Nisa Ulhusna, Rahmadani, Mira Meilisa — Tue, 03 Mar 2026 00:00:00 +0000

Job Satisfaction is a essential factor influencing employee performance, commitment, and organizational sustainability. Low levels of Job Resources and suboptimal Leadership Quality are common causes of decreased job satisfaction across various institutions. This study aims to analyze the effect of job resources and leadership quality on Jjob Satisfaction using the Structural Equation Modeling (SEM) method. The research data were obtained from a Likert-scale survey (1-8) consisting of three latent variabless and their respective indicators, and wer analyzed through Confirmatory Factor Analysis (CFA) and Structural Model assesment. The result of the CFA indicate that all indicators meet the criteria for validity and reliability, with factor loadings above 0.50, a Composite Reliability (CR) value of 0.9667, and an Average Variance Extracted (AVE) value of 0.6769. the Goodness of Fit evaluation shows that the final model is highly acceptable, as reflected by a low Chi-square/df value, RMSEA = 0.005, and CFI, TLI, GFI, and NFI value of 1.000. the Structural analysis further demonstrates that Job Resources have a positive and significant impact on Job Satisfaction. Simultaneously, both variables contribute significantly to explaining variations in Job Satisfaction. This study highlights that enhancing Job Resources and improving Leadership Quality are crucial strategies to strengthen employee Job Satisfaction. The findings provide empirical insight that can assist organizations in developing more effective and sustainable human resource management policies

Fuzzy Time Series Singh Method for Forecasting Tourist Arrivals at Kinantan Wildlife and Cultural Park Bukittinggi

Olivin Adelia Huqmi, Fadhilah Fitri, Tessy Octavia Mukhti — Mon, 16 Mar 2026 00:00:00 +0000

Tourism is a key sector in regional development, contributing to economic growth, job creation, and cultural preservation. In Bukittinggi, West Sumatra, the Kinantan Wildlife and Cultural Park (TMSBK) is a major tourist destination, known for its historical and educational value. Tourist visits to TMSBK show fluctuating trends influenced by seasonal factors, socio-economic conditions, and national or global events. These dynamics make accurate forecasting essential for effective tourism planning and management. This study aims to forecast monthly tourist visits to TMSBK using the Fuzzy Time Series (FTS) Singh method, which is suitable for uncertain and fluctuating time series data. The research used historical visitor data from 2021 to 2024 obtained from the Central Bureau of Statistics. The forecasting process included defining the universe of discourse, forming class intervals, fuzzifying historical data, establishing fuzzy logical relationships (FLR), and generating forecasts. The accuracy of the forecasts was measured using Mean Absolute Percentage Error (MAPE), with a result of 19.8%, indicating good predictive performance. The results show that the FTS Singh method successfully follows the fluctuation pattern of actual visitor data. This method provides valuable insights for destination managers in planning operations, promotional efforts, and service improvements. Therefore, the FTS Singh method can be considered a reliable tool to support sustainable tourism development and decision-making in Bukittinggi.

Application of Fuzzy Time Series Cheng in Forecasting Bukittinggi Consumer Price Index

Afifah Nabilah, Fadhilah Fitri, Yenni Kurniawati — Mon, 16 Mar 2026 00:00:00 +0000

The Consumer Price Index (CPI) is one of the main indicators used to measure inflation and assess the public’s purchasing power. Based on CPI monitoring in March 2025, Bukittinggi City recorded the highest year-on-year (y-o-y) inflation rate in West Sumatra at 0.50 percent, with a CPI of 106.99. This indicates significant price fluctuations, which require careful analysis and forecasting to support regional economic policymaking. This study aims to forecast the CPI of Bukittinggi City for April 2025 using the Fuzzy Time Series (FTS) Cheng method. The data used consists of monthly CPI values from January 2020 to March 2025, totaling 63 observations, obtained from the official website of Statistics Indonesia (BPS). The forecasting result using the FTS Cheng method for April 2025 shows a CPI value of 106.19. To evaluate the model's accuracy, the Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE) were employed, yielding values of 0.82% and 0.90%, respectively. These values fall into the “very good” category based on standard forecasting accuracy criteria. The FTS Cheng method was selected due to its ability to accommodate data fluctuations and provide weighted relationships between fuzzy intervals, thus enhancing forecasting accuracy in dynamic economic conditions. However, this study is limited to univariate data and does not compare the FTS Cheng method with other forecasting models. This research provides valuable insights for local governments in designing effective economic strategies based on reliable predictive models.

Spatial Autoregressive Model to Factors Poverty Gap Index in West Java 2023

Rahmat Kurniawan, Figo Rahmatullah, Fauzan Gustiandra, Tessy Octavia Mukhti — Mon, 16 Mar 2026 00:00:00 +0000

Spatial analysis is the analysis of data with spatial effects. The spatial autoregressive is used when the effect of the dependent variable at one location is influenced by the value of the dependent variable at nearby or neighboring locations. The spatial autoregressive model is more appropriate to model the factors influencing the poverty depth index in West Java in 2023. Based on the Spatial Autoregressive modeling, the variables that influence the Poverty Depth Index in West Java are Population Density, Open Unemployment Rate, and economic growth. The SAR modeling produces a higher coefficient of determination compared to the linear model, which is 68.88% with an AIC value of 18.6149.

Grouping of Provinces in Indonesia Based on Quality Education Indicators in 2025 Using the Self-Organizing Maps Method

Dinda Putri Adilla, Tessy Octavia Mukhti — Mon, 16 Mar 2026 00:00:00 +0000

Access to quality education plays an essential role in improving people’s welfare and supporting sustainable development. As a fundamental component of social progress, quality education is not limited to academic attainment but also involves the development of skills, values, and character needed for meaningful participation in society. This study seeks to identify patterns and disparities in education quality among provinces by grouping regions based on multiple educational indicators. The indicators analyzed include Average Years of Schooling, Literacy Rate, Access to Information and Communication Technology, Gross Enrollment Rate, Net Enrollment Rate, and Teacher Qualifications. The data were examined using descriptive statistics, data visualization, and normalization, followed by clustering through the Self-Organizing Map (SOM) method as an unsupervised learning approach in data mining. Two clusters were formed to represent provinces with relatively higher and lower levels of educational quality. Cluster validity was assessed using internal validation measures, namely the Connectivity Index, Silhouette Index, and Dunn Index. The findings reveal that most basic education and literacy indicators show relatively favorable conditions; however, disparities remain evident in average years of schooling, ICT access, and participation in secondary and higher education. The clustering results indicate that 35 provinces fall into the group with relatively higher education quality, while 3 provinces are classified in the lower category. These results suggest that although the overall condition of education is relatively good, regional inequality in educational outcomes persists and requires targeted policy interventions to promote more balanced and inclusive development.

Comparison of Agglomerative Hierarchical Clustering Methods for Grouping Indonesian Provinces Based on Community Literacy Development Index

Olga Afrilly Putri, Bunga Nafandra, Zamahsary Martha — Mon, 16 Mar 2026 00:00:00 +0000

Community literacy development is one of the important indicators in improving the quality of human resources in Indonesia. This study aims to group provinces in Indonesia based on the Community Literacy Development Index by considering the equity of library services, the adequacy of library collections, and the level of community visits per day. The method used is agglomerative hierarchical cluster analysis. Before grouping, the data is standardized to overcome differences in units and scales between variables. The selection of the best cluster method is done using the cophenetic correlation coefficient, while the determination of the optimal number of clusters uses the silhouette method. The results of the analysis show that the Average Linkage method is the most optimal hierarchical cluster method with the best number of clusters being four clusters. Each cluster has different characteristics, reflecting variations in community literacy levels, service equity, collection adequacy, and library visit intensity between provinces. These findings indicate disparities in community literacy development between regions in Indonesia. Therefore, the results of this study are expected to serve as a basis for consideration in formulating more effective and targeted literacy and library development policies.