JURNAL ILMIAH MATEMATIKA DAN TERAPAN

Spatial Statistical Analysis for Poverty Mapping Using Machine Learning

Agung Yuliyanto Nugroho — 2025-12-31

Poverty is a multidimensional problem influenced not only by economic factors but also by spatial dimensions such as geographic location, accessibility, and environmental characteristics. This study aims to analyze spatial patterns of poverty and develop a poverty prediction model using a geospatial data-based machine learning approach. The data used comes from a combination of open sources such as the Central Statistics Agency (BPS), Landsat satellite imagery, and regional infrastructure data. The methods used include spatial autocorrelation analysis (Moran's I) to identify poverty clustering patterns, Local Indicators of Spatial Association (LISA) to detect poverty hotspots, and Random Forest and Gradient Boosting models to predict poverty levels based on environmental, social, and economic variables. The results show that poverty has a significant spatial pattern, where areas with high poverty rates tend to cluster in areas with low infrastructure access and high population density. The machine learning model demonstrated better prediction accuracy than the traditional linear regression approach, with an R² value reaching 0.87 and a lower prediction error rate (RMSE). These findings emphasize the importance of integrating spatial analysis and machine learning technology in understanding the dynamics of poverty geographically. This research contributes to the development of spatial data analysis methods in the context of public policy, particularly in supporting more targeted poverty alleviation intervention planning. The mapping results can serve as a basis for local governments in identifying priority areas, allocating resources, and designing data-driven development policies. Thus, this approach offers an innovative solution towards more efficient and evidence-based decision-making in poverty alleviation in Indonesia.

Analysis of Stability of Lubricating Oil Discharge Using the Individual Moving Range (IM-R) Chart Method

Ayu Nur Fithri — 2025-12-31

The stability of lubricant oil expenditure at PetroChina International Jabung Ltd with the aim of preventing dead stock. Dead stock is defined as material stock that has not been used or issued for more than five years. The existence of dead stock can lead to excess materials, which results in wasted costs. Specifically, if PetroChina experiences excess materials, the funds used to purchase lubricating oil will not be reimbursed by the state. The data used is oil lubricating production data for the period January 2022-December 2024 each month with a total of 36 observations. To analyze oil lubricating production stability, an Individual Moving Range (IM-R) chart is used, which is a Statistical Quality Control (SQC). The analysis results show several data points that are outside the control limits, indicating special cause variations in the oil lubricating production process. These uncontrolled points indicate that the process is not yet fully stable and can be influenced by factors outside of normal variations. The results of the study provide recommendations in the form of further investigation into the surge in oil lubricating production and optimizing demand planning through ROP/ROQ and SOQ to make oil lubricating production more consistent and avoid the risk of dead stock.

Application of the KMeans Clustering Algorithm in E-Commerce Transaction Pattern Analysis

Agung Yuliyanto Nugroho — 2025-12-31

In the era of digital transformation, e-commerce platforms have become a major driver of economic activity, generating vast amounts of transaction data every day. Analyzing these data can provide valuable insights into customer behavior, purchasing trends, and business performance. This study aims to apply the K-Means clustering algorithm to identify and analyze transaction patterns in e-commerce systems. The research focuses on developing an efficient data-driven approach to segment customers based on their transactional attributes, such as purchase frequency, transaction value, and product category preferences. The methodology involves several stages: data preprocessing, including cleaning and normalization; feature selection based on relevant transactional indicators; and the application of the K-Means clustering algorithm to group customers into clusters with similar characteristics. The Elbow Method was used to determine the optimal number of clusters. Data were processed using the Python programming language and libraries such as Scikit-learn and Pandas. The results reveal that K-Means effectively segments e-commerce customers into distinct groups that reflect their purchasing patterns—ranging from high-value loyal customers to occasional buyers. Each cluster presents unique behavioral profiles that can be interpreted for targeted marketing strategies. The clustering outcome provides useful insights for customer relationship management (CRM), inventory optimization, and personalized product recommendations. In conclusion, the application of the K-Means algorithm demonstrates significant potential in uncovering hidden patterns within large-scale e-commerce transaction data. The findings support the use of mathematical and computational models in improving decision-making processes in digital commerce. Future research is recommended to enhance cluster accuracy by integrating hybrid algorithms or deep learning-based segmentation approaches.

Utilization of Data Analysis-based Educational Technology to Improve Learning Participation of Students With Special Needs at SDN Cipondoh 03

Siska Damayanti — 2025-12-31

This study aims to analyze the effectiveness of data-driven educational technology in improving learning participation among special-needs students in inclusive primary schools. Many inclusive institutions in Indonesia still face challenges in enhancing active participation among students with special educational needs due to limited adaptive instructional strategies and minimal use of learning analytics. The purpose of this research is to investigate how the integration of data-based educational technology supports engagement, interaction, and inclusiveness in the classroom.

This study employed a Classroom Action Research (CAR) method conducted in two cycles at SDN Cipondoh 03, Tangerang City. Participants consisted of 20 special-needs students and one classroom teacher. Data were collected through classroom observation, interviews, documentation, and analysis of digital learning logs. Quantitative data measured participation rates, while qualitative data examined behavioral and motivational changes. Analysis combined descriptive statistics and reflective interpretation of teacher observations.

Results show a significant improvement in student participation—from 45% in the pre-cycle to 68% in the first cycle, and 82% in the second cycle. The increase is supported by the use of learning analytics dashboards that provide personalized feedback and visual progress reports for students. These findings confirm that data-driven educational technology effectively enhances participation and engagement of special-needs students by fostering adaptive instruction and evidence-based teaching.