Pangestu, Dito Ananda (2024) Prediksi Gempa Daerah Sulawesi Menggunakan Time Series Analysis. Other thesis, Universitas Islam Riau.

Text Dito Ananda Pangestu.pdf - Submitted Version Restricted to Registered users only Download (7MB) | Request a copy

Abstract

Earthquakes are one of the natural disasters that often occur, every time an earthquake occurs, there will definitely be damage. An earthquake is a vibration that occurs on the earth's surface due to the sudden release of energy from within which creates seismic waves. In Indonesia, Sulawesi is one of the regions in Indonesia that frequently experiences earthquakes and high seismic activity. This is due to its geographical location in the meeting zone of three main tectonic plates: the Indo-Australian Plate, the Eurasian Plate, and the Philippine Plate. Sulawesi still does not have an Earthquake Early Warnings system, which is a very fast earthquake detection system used in Japan. in this system is data describing earthquake predictions to be able to carry them out. Therefore, earthquake predictions use machine learning algorithms, namely Facebook Prophet and Long Short Term Memory. The dataset used was taken from data, namely Sulawesi earthquake data from BMKG. The best prediction results were the Long Short Term Memory algorithm which got an MAE value of 0.58777 with an RSME of 0.73675. The Facebook Prophet algorithm got an MAE value of 0.59018 with an RSME of 0.75647. In this research, the algorithm that had the best results was the Long Short Term Memory algorithm in predicting Time Series earthquakes in the Sulawesi area.

Actions (login required)

View Item