PENGEMBANGAN MODEL SUPPORT VECTOR MACHINE (SVM) DAN GEOSPATIAL ARTIFICIAL INTELLIGENCE (GEOAI) UNTUK PREDIKSI WILAYAH ENDEMIS DEMAM BERDARAH

MEILENI, HETTY and Ermatita, Ermatita and Abdiansah, Abdiansah and Husni, Nyayu Latifah (2025) PENGEMBANGAN MODEL SUPPORT VECTOR MACHINE (SVM) DAN GEOSPATIAL ARTIFICIAL INTELLIGENCE (GEOAI) UNTUK PREDIKSI WILAYAH ENDEMIS DEMAM BERDARAH. Doctoral thesis, Sriwijaya University.

Preview	Image RAMA_21001_03013622227008_cover.jpg - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (172kB) | Preview
	Text RAMA_21001_03013622227008.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (10MB) | Request a copy
	Text RAMA_21001_03013622227008_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (27MB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (2MB)
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (526kB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (874kB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (258kB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (235kB) | Request a copy
	Text RAMA_21001_03013622227008_0013096707_0001108401_ 0003057603_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (6MB) | Request a copy

Abstract

This study uses Support Vector Machine (SVM), Support Vector Regression (SVR), Geographically Weighted Regression (GWR), and Geospatial Artificial Intelligence (GeoAI) to make a model that can predict where Dengue Fever (DF) is most likely to happen. The study utilizes DHF epidemiological data from 2017 to 2023, which included the number of DF cases, population density, climatic variables (temperature, rainfall, and humidity), and spatial data from each city and sub-city. This study has three main parts: (1) using SVM to sort areas into endemic and non-endemic groups; (2) using SVR to guess the number of DF cases; and (3) using GWR to look at spatial relationships. We checked the model's accuracy with Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and GeoAI-based spatial mapping. With a 98.3% accuracy rate, a 96.6% recall rate, and an F1 score of 99.44%, SVM was able to correctly identify DF-endemic areas 99.12% of the time. The linear kernel in the SVR model gave an MAE of 0.096, an MSE of 0.009, an RMSE of 0.097, and a MAPE of 30.79%, showing that it was very good at predicting the future. The Moran's I value of 0.00131 and the AIC value of 150.04857 show that GWR with the Bi-Square kernel was better than the RBF method at capturing the spatial patterns of DF. This study is a big step toward making a better early warning system for DF control. This will help with making strategic decisions like where to send medical staff, how to use resources, and how to target public awareness campaigns. Overall, this study shows that combining SVM, SVR, and GeoAI can make a good predictive model for mapping and analyzing areas where dengue fever is common. This improves the accuracy of predictions and builds a strong base for South Sumatra Province's plans to control infectious diseases. Keywords: Dengue Fever (DF), Support Vector Machine (SVM), Support Vector Regression (SVR), Geographically Weighted Regression (GWR), GeoAI, endemic area prediction, spatial modeling.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	Dengue Fever (DF), Support Vector Machine (SVM), Support Vector Regression (SVR), Geographically Weighted Regression (GWR), GeoAI, endemic area prediction, spatial modeling
Subjects:	Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions:	03-Faculty of Engineering > 21001-Engineering Science (S3)
Depositing User:	Hetty Meileni
Date Deposited:	08 Apr 2025 01:54
Last Modified:	08 Apr 2025 01:54
URI:	http://repository.unsri.ac.id/id/eprint/169899

Actions (login required)

View Item