Prediction for Dengue Fever in Indonesia Using Neural Network and Regression Method
T. Henny Febriana Harumy, Huah Yong Chan, Gian Chand Sodhy · Journal of Physics Conference Series · 2020
Abstract Dengue fever is the most hurriedly diffused mosquito-borne viral disease in the world. More than 33% of the total population in the world is under risk. Currently the prediction of dengue can save a person ‘s life by alerting them to take proper diagnosis and care. The Objectives of this research is (1) To Predict The area with the most potential to suspend dengue fever In Indonesia, And (2) to Predict dengue fever cases. (3). To analyz how many percent the effect factor of dengue fever. There are many ways to predict one of them is Regression and Deep learning Approach. Reseacher tried to analyz the most accurate such as Regression Multyplied, Neural Network, and Sensitivity Analysis. Set of data have been used is timeseries from 1997—2017. The Variable has been used for this research is Humadity, Temperature, Wind, Airpressure, Rainfall Index, income, sunlight, Population density, and output is cases. The Result of this research is first The area with the most potential to suspend Dengue fever In Indonesia 2019 is Jambi, Lampung, Bangka Belitung, West Sumatera, Central java with average Accuracy 87,16%. The Prediction dengue fever cases 2019 is 80233 Cases with accuracy 87,16%. The Third All variable ( X1 s.d X8) have been to effect to Partial and Simultaneous to (Y) in the amount of 0.16 (16 % ) with a significance level of 0.001 (99%). While the remaining 100% - 16% = 84% is influenced by other variables outside of this research.