Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Dinasti Information and Technology
Muhammad Nauvaldi Akbar A
Department of Informatics, Faculty of Engineering, Universitas Muhammadiyah Surabaya, Indonesia
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Dengue Hemorrhagic Fever Case Prediction Using Climate and Social Media Data with Support Vector Regression Erie Kresna Andana; Muhammad Nauvaldi Akbar A; Sri Amaliah Mandati
Dinasti Information and Technology Vol. 4 No. 1 (2026): Dinasti Information and Technology (July - September 2026)
Publisher : Dinasti Research & Yayasan Dharma Indonesia Tercinta (DINASTI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.38035/dit.v4i1.3884

Abstract

Dengue hemorrhagic fever (DHF) remains a recurring public health problem in Surabaya, with transmission influenced by climatic variability and potentially reflected in social media activity. This study aims to develop an informatics-based model for predicting annual DHF cases by integrating climate and Twitter/X data and comparing Support Vector Regression (SVR) with Ridge Regression. The dataset covers 2010–2024 and combines annual DHF case records, annual and seasonal climate indices from Juanda and Tanjung Perak stations, and 320 Indonesian-language tweets related to DHF in Surabaya. Social media data were transformed into annual tweet volume, negative-sentiment tweet volume, and mean sentiment score, while predictors were aligned using a one-year time lag. Model performance was evaluated using Leave-One-Out Cross-Validation for 2011–2024 based on MAE, RMSE, and R². SVR achieved the best performance with an MAE of 386.31 cases, RMSE of 556.86 cases, and R² of 0.0121, outperforming Ridge Regression with an MAE of 601.31, RMSE of 727.19, and R² of −0.6846. These findings indicate that SVR better captures nonlinear relationships in multisource dengue data, although explanatory power remains limited by few annual observations and high feature dimensionality. This study demonstrates the potential of multisource data integration for data-driven dengue early warning systems.
Co-Authors Erie Kresna Andana Sri Amaliah Mandati