PREDICTION OF DENGUE HEMORRHAGIC FEVER (DHF) CASE DISTRIBUTION USING LONG SHORT-TERM MEMORY (LTSM) IN SURABAYA CITY

Abstract

Background: The incidence of dengue fever in Indonesia in 2023 reached 41.4 per 100,000 population, which is classified as high according to the Indonesian Ministry of Health (with a target IR < 10/100,000 population). Surabaya City is one of the areas with a significant increase in DHF cases, and has the highest population density in East Java. Object: This study aims to produce a predictive model that can identify trends in dengue incidence using the Long Short-Term Memory (LSTM) method in Surabaya City. Methods: This research is a quantitative study with a retrospective approach using secondary data from 2020 to 2024. The variables analysed included rainfall, humidity, temperature, larva-free rate, and the number of dengue cases. Results: The results of the analysis showed a significant relationship between rainfall (r = -0.195; p = 0.033), temperature (r = 0.276; p = 0.002), and ABJ (r = -0.590; p = 0.000) with the incidence of DHF. Humidity showed no significant association (r = 0.012; p = 0.894). Conclusion: It is recommended that strengthening the 3M Plus programme and integrating an LSTM-based early warning system are carried out as preventive measures in public health strategies.