Prediksi Risiko Demam Berdarah Menggunakan VMD-BILSTM Pada Kabupaten Malang

Sava, Arvin Azmi (2025) Prediksi Risiko Demam Berdarah Menggunakan VMD-BILSTM Pada Kabupaten Malang. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026211097-Undergraduate_Thesis.pdf Restricted to Repository staff only until 1 April 2027. Download (9MB) | Request a copy

Abstract

Demam berdarah dengue (DBD) merupakan penyakit yang ditularkan melalui gigitan nyamuk Aedes Aegypti. Peningkatan jumlah kasus DBD menunjukkan bahwa penyakit ini menjadi ancaman kesehatan masyarakat yang serius dan membutuhkan perhatian global. Pola penyebaran DBD yang cepat dan meluas menyebabkan dampak yang signifikan bagi masyarakat, terutama di daerah dengan sistem kesehatan yang terbatas. Oleh karena itu, penelitian tugas akhir ini bertujuan untuk melakukan prediksi risiko wabah demam berdarah dengan menggunakan pendekatan Variational Mode Decomposition -Bidirectional Long Short-Term Memory (VMD-BiLSTM). Hasil penelitian menunjukkan bahwa model VMD-BiLSTM mampu memprediksi angka insiden DBD dengan baik, lebih unggul daripada model konvensional BiGRU dan BiLSTM, serta mampu meningkatkan akurasi hasil prediksi hingga 39% dibandingkan BiGRU dan BiLSTM ditinjau dari RMSE. Oleh karena itu, dapat disimpulkan proses dekomposisi VMD berpengaruh dalam meningkatkan akurasi model prediksi. Penambahan variabel sosial media juga berpengaruh dalam meningkatan akurasi prediksi walaupun efektivitasnya masih terbatas. Hasil akurasi terbaik yang mampu dihasilkan VMD-BiLSTM dalam melakukan prediksi angka insiden demam berdarah terjadi pada Kecamatan Jabung dengan RMSE sebesar 1,12, SMAPE sebesar 27,43, dan MAE sebesar 0,75.
==================================================================================================================================
Dengue hemorrhagic fever (DHF) is a disease transmitted through the bite of the Aedes Aegypti mosquito. The increasing number of DHF cases indicates that the disease is becoming a serious public health threat and requires global attention. The rapid and widespread pattern of dengue spread causes a significant impact on society, especially in areas with limited health systems. Therefore, this final project research aims to predict the risk of dengue fever outbreaks using the Variational Mode Decomposition-Bidirectional Long Short-Term Memory (VMDBiLSTM) approach. The results demonstrate that the VMD-BiLSTM model effectively predicts dengue incidence rates, achieving better performance compared to conventional BiGRU and biLSTM models Moreover, it is capable of improving prediction accuracy by up to 39% compared to BiGRU and BiLSTM, as evaluated using RMSE. Therefore, it can be concluded that the VMD decomposition process contributes to improving the accuracy of the prediction model. The addition of social media variables also has an effect in improving prediction accuracy although its effectiveness is still limited. The highest accuracy achieved by the VMD-BiLSTM model in predicting dengue fever incidence was observed in Jabung District, with an RMSE of 1.12, SMAPE of 27.43, and MAE of 0.75.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	dengue fever, risk prediction, Variational Mode Decomposition, BiLSTM, demam berdarah, prediksi risiko, Variational Mode Decomposition, BiLSTM.
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Arvin Azmi Sava
Date Deposited:	03 Feb 2025 07:50
Last Modified:	03 Feb 2025 07:50
URI:	http://repository.its.ac.id/id/eprint/117961

Actions (login required)

View Item