Penerapan Algorimta Backpropagation Untuk Prakiraan Cuaca Harian Dibandingkan Dengan Support Vector Machine dan Logistic Regression

Ayu Zulfiani, Chairani Fauzi

JURNAL MEDIA INFORMATIKA BUDIDARMA · 23 July 2023

In one sentence. Backpropagation neural network predicts classified daily rainfall at a Lampung station, outperforming SVM and logistic regression on light-rain recall.

Abstract

To anticipate the impacts caused by extreme weather, the Meteorology, Climatology, and Geophysics Agency (BMKG) issues weather forecasts so that the community can be prepared when such extreme weather occurs. The application of Artificial Neural Network (ANN) techniques in weather forecasting significantly enhances the ability to explore vast amounts of big data in obtaining the necessary information, serving as a reliable assistant for forecasting and policymaking. The data used in this study consists of weather elements such as pressure, air temperature, humidity, wind direction and speed, as well as rainfall, obtained from the Radin Inten II Lampung Meteorological Station. The observational data has a data density per hour, spanning a period of 5 years from January 1, 2018, to December 31, 2022. The method employed in this research is Backpropagation Neural Network (BPNN). The research results indicate that BPNN can effectively predict classified rainfall compared to other methods, within recall value when slight rain 0.68, moderate rain 0.17, and heavy rain 0.03, meanwhile Support Vector Machine (SVM) and Logistic Regression (LR) method can predict only slight rain with recall value when slight rain is 0.51 and 0.47.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.