JURNAL SINTAK

Identification Of Disparities In Educational Facilities Among Indonesian Provinces Using K-Means Clustering

Ananda Desilia Putri — 2026-03-25

Education is one of the indicators used to measure a country's progress, due to its important role in producing quality human resources. Good education is certainly supported by the availability of adequate educational facilities. However, Indonesia's diverse geography poses a challenge in terms of equal access to education. Provinces classified as 3T regions face a shortage of educational facilities and teaching staff. This study aims to group provinces in Indonesia based on educational facility indicators for the 2023/2024 academic year using K-Means Clustering Analysis. The data used covers 38 provinces with 18 educational facility indicators, which were analyzed after data pre-processing. The results of this study obtained 3 clusters, where the first cluster consisted of 1 province with poor access and infrastructure conditions, the second cluster consisted of 17 provinces with fairly good access and infrastructure conditions, and the third cluster consisted of 20 provinces with very good access and infrastructure conditions. The clustering results from this study are expected to serve as a reference for the formulation of policies on the equitable distribution of educational facilities and the determination of development priorities in the education sector in Indonesia.

Identification of Indonesian Provinces Based on Socioeconomic Indicators in 2024 Using K-Means

Erika Putri Permatasari — 2026-03-25

Poverty remains a persistent development challenge in Indonesia, characterized by substantial disparities across regions. Differences in social and economic conditions among provinces highlight the need for a comprehensive regional classification to support the formulation of targeted development policies. This study aims to classify Indonesian provinces based on their poverty and development characteristics. The data used are secondary data for the year 2024 obtained from Statistics Indonesia (Badan Pusat Statistik), with 38 provinces as the units of analysis. The variables include the poverty rate, Human Development Index (HDI), Open Unemployment Rate (OUR), and gross regional domestic product (GRDP) per capita. The analytical method employed is K-Means clustering, with variables standardized using Z-Scores. The optimal number of clusters was determined using the Elbow method and confirmed by the Silhouette Score. The results indicate that Indonesian provinces can be grouped into four clusters with distinct social and economic characteristics. Each cluster reflects different levels of poverty, human development quality, and labor market conditions. These findings emphasize that poverty in Indonesia is a multidimensional issue, underscoring the need for development and poverty alleviation policies that are tailored to the specific characteristics of each cluster.

Segmentation Of Educational Quality In Indonesian Provinces Based On K-Means Clustering

Baiq Jasmin Sabhira Safwa V.R — 2026-03-25

The quality of education in Indonesia still exhibits disparities among provinces, reflecting differences in educational attainment and access. This study aims to segment the quality of education across Indonesian provinces based on the similarity of educational characteristics using the K-Means Clustering method. The data used consist of provincial-level education data that have undergone outlier detection and standardization to ensure comparability across variables. K-Means Clustering analysis was performed by forming three clusters representing provinces with low, medium, and high levels of educational quality. The clustering results indicate that most provinces fall into the medium education quality cluster, while a smaller number of provinces remain in the low education quality cluster. These findings demonstrate that the K-Means Clustering method is able to provide a clear representation of segmentation patterns and disparities in educational quality across Indonesian provinces and can serve as a basis for supporting more targeted and equity-oriented education policy formulation.

Keywords: education; quality; K-Means; clustering; provinces

Correlation Analysis Between Material Thickness and Welding Length on The Completion Time of Vessel Product in The Static Mixer Project

Salsabila Bahri — 2026-03-25

This study aims to analyze the relationship between fabrication technical parameters (material thickness and welding length) and the completion time of static mixer vessel products. The research sample consisted of 30 product units, with data collected retrospectively from the project documentation of PT. NOV Profab for the period December 2024 until July 2025. The method used was quantitative correlational. The Shapiro-Wilk normality test indicated that the data was not normally distributed (p < 0.05), therefore correlation analysis was performed using the non parametric Spearman’s Rank test. The results show that the relationship between material thickness and completion time is very weak and not significant (r = 0.126 and p = 0.507). Similarly, the relationship between welding length and completion time is weak and not significant (r = 0.301 and p = 0.106). In conclusion, material thickness and welding length are not proven to have a statistically significant relationship with the completion duration of static mixer products. This finding implies that project time estimation requires consideration of factors other that these technical parameters.

Cluster Analysis and Discriminant Analysis for Grouping Provinces Based on Factors Affecting Poverty Levels in Indonesia 2018-2020

Kariyam — 2026-03-25

Poverty is a condition that occurs due to the inability of a person or group to meet the minimum basic needs, such as food, clothing, health, housing, and education, which are necessary to maintain survival. The poverty level of an area is influenced by various factors, including the Open Unemployment Rate (TPT), the Provincial Minimum Wage (UMP), and the Human Development Index (IPM). This research aims to group provinces in Indonesia based on factors that affect poverty and determine the discriminatory function of the group formed. The analysis method used is cluster analysis to group provinces into several poverty level groups and discriminatory analysis to form a separating function between the groups. The results of cluster analysis show the formation of three groups, namely the group with the highest poverty level consisting of 7 provinces, the group with moderate poverty level consisting of 8 provinces, and the group with the lowest poverty level which includes other provinces. Furthermore, discriminant analysis produces a discriminant function that is able to distinguish between poverty levels quite well. The results of this research are expected to be considered by the government in formulating poverty alleviation policies that are more on target

Forecasting the Consumption of Welding Consumables Using Markov Chain Models for Inventory Optimization at PT Buana Cipta Mandala

Muhammad Reza Rafella Palevi — 2026-03-25

This study aims to analyze the monthly withdrawal patterns of welding consumables (welding cup, black lens, clear lens) and develop a forecasting model to support inventory policy optimization at PT Buana Cipta Mandala, Batam. Employing a quantitative case study approach with time series data from Januari to August 2025. Withdawal volume data was categorized into three states (low, medium, high). The Markov chain model was constructed by calculating transition frequency matrices, transition probability matrices, and steady-state probabilities for each item. Preliminary descriptive statistical analysis was conducted to understand data characteristics. The findings reveal distinct transition patterns. The welding cup exhibits a rapid cycle dynamic with a steady-state probability 0.286 for low, 0.286 for medium, and 0.428 for high state, indicating a long term dominance of the high state. Conversely, the welding lenses have a transition matrix where the low state acts as an absorbing state, with a steady-state probability 1 for low, and 0 for medium and high state, predicting a convergence of demand to a low level. The resulting model recommends differentiated inventory strategies. A moderate to high stock policy with sufficient safety stock for welding cups, and a lean inventory policy based on base demand for welding lenses. The application of this Markov chain model provides a quantitative foundation for more precise procurement decision making, reducing the risks of stockout and overstocking, thereby supporting supply chain efficiency and shipyard operations.