Rahmasari, Hazelita Dwi
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Flood Risk Clustering Based on SARIMA Rainfall Prediction and Regional Mapping in Central Java Maulidiyah, Wildatul; Rahmasari, Hazelita Dwi; Notodiputro, Khairil Anwar; Angraini, Yenni; Mualifah, Laily Nissa Atul 
Journal of Mathematics, Computations and Statistics Vol. 8 No. 2 (2025): Volume 08 Nomor 02 (Oktober 2025)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/jmathcos.v8i2.8632

Abstract

High and spatially uneven rainfall is a major contributing factor to flooding in tropical regions such as Indonesia, including Central Java Province. This study aims to classify regions based on rainfall patterns using the Dynamic Time Warping (DTW) method and hierarchical clustering, followed by rainfall forecasting for each cluster using the SARIMA model. The dataset comprises monthly rainfall records from 2017 to 2023 across 35 regencies and cities in Central Java. The clustering process identified three distinct groups with low, medium, and high rainfall intensity. Evaluation results indicated that the single linkage models for each cluster were SARIMA(0,0,2)(0,1,0)[12] with a MAPE of 27% (Cluster 1), SARIMA(0,1,2)(0,1,1)[12] with a MAPE of 9.4% (Cluster 2), and SARIMA(1,0,0)(1,1,0)[12] with a MAPE of 9.97% (Cluster 3). These findings provide a robust spatio-temporal basis for supporting flood risk mitigation strategies based on rainfall prediction in Central Java.
PERBANDINGAN K-MEANS DAN K-MEDOIDS DALAM PENGELOMPOKKAN KOMODITAS EKSPOR INDUSTRI DI INDONESIA Arifa, Panji Lokajaya; Rahmasari, Hazelita Dwi; Aimandiga, Carlya Agmis; Fitrianto, Anwar; Yudhianto, Rachmat Bintang
MUST: Journal of Mathematics Education, Science and Technology Vol 10 No 2 (2025): DECEMBER
Publisher : Universitas Muhammadiyah Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30651/must.v10i2.28661

Abstract

International trade plays a crucial role in Indonesia's economic growth, particularly through industrial commodity exports. However, its heavy dependence on a few key commodities makes it vulnerable to global market fluctuations. This study aims to explore trends in industrial commodity export values ​​and compare the performance of cluster methods in grouping commodities based on their value patterns. The research data used are monthly export values ​​from 2022 to mid-2025, sourced from the Central Statistics Agency (BPS). The analytical methods used include trend exploration and cluster analysis with K-Means and K-Medoids using Dynamic Time Warping (DTW) distance. The results of the export value trend exploration indicate that palm oil dominates industrial export value, while other commodities tend to have stable patterns at medium to low values. Evaluation of clustering results using K-Means and K-Medoids each obtained 3 clusters indicating that K-Medoids provided the best performance by obtaining a Silhouette Score of 0.1577 and a Davies-Bouldin Index (DBI) of 1.7990. This value is better than K-Means which obtained a Silhouette Score of 0.1493 and a DBI of 2.3037 indicating that the method is less than optimal in separating clusters. This finding explains that K-Medoids is more robust against outliers and is able to provide more representative groupings. So it can provide a deeper understanding of commodity grouping patterns and contribute to providing export policy recommendations to reduce dependence on primary commodities and increase the export competitiveness of Indonesian industrial products.