PENERAPAN HYBRID SEASONAL AUTOREGRESSIVE INTEGRATED MOVING AVERAGE – LONG SHORT TERM MEMORY (SARIMA-LSTM) DALAM MERAMALKAN CURAH HUJAN DI BOGOR

Annisa Nur Azizah, - (2023) PENERAPAN HYBRID SEASONAL AUTOREGRESSIVE INTEGRATED MOVING AVERAGE – LONG SHORT TERM MEMORY (SARIMA-LSTM) DALAM MERAMALKAN CURAH HUJAN DI BOGOR. S1 thesis, Universitas Pendidikan Indonesia.

	Text S_MAT_1900383_Title.pdf Download (376kB)
	Text S_MAT_1900383_Title.pdf Download (376kB)
	Text S_MAT_1900383_Chapter I.pdf Download (312kB)
	Text S_MAT_1900383_Chapter II.pdf Restricted to Staf Perpustakaan Download (570kB)
	Text S_MAT_1900383_Chapter III.pdf Download (334kB)
	Text S_MAT_1900383_Chapter IV.pdf Restricted to Staf Perpustakaan Download (561kB)
	Text S_MAT_1900383_Chapter V.pdf Download (169kB)
	Text S_MAT_1900383_Appendix.pdf Restricted to Staf Perpustakaan Download (316kB)

Abstract

Curah hujan merupakan salah satu faktor yang paling berpengaruh bagi kehidupan manusia. Curah hujan yang tinggi menyebabkan terjadinya bencana alam seperti banjir dan longsor. Peramalan curah hujan sangatlah penting untuk mempermudah mengambil keputusan dalam menghadapi keadaan jika terjadi curah hujan tinggi. Model yang sering digunakan dalam peramalan adalah model SARIMA (Seasonal Autoregressive Integrated Moving Average) karena dianggap cukup sederhana. Namun, model SARIMA memiliki kelemahan yaitu hanya mampu menangkap pola data linear. Pada kenyataanya, tidak semua data bersifat linear termasuk curah hujan. Pada penelitian ini dilakukan pemodelan hybrid SARIMA dengan jaringan syaraf tiruan LSTM (Long Short Term Memory) untuk mengatasi permasalahan data non linear. Pemodelan hybrid diharapkan dapat meningkatkan akurasi hasil peramalan. Hasil akhir menunjukkan jika model hybrid SARIMA-LSTM memiliki hasil dan nilai ketepatan yang lebih baik dibandingkan dengan model SARIMA. Rainfall is one of the most influential factors for human life. High rainfall causes natural disasters such as floods and landslides. Forecasting rainfall is very important to make it easier to make decisions in dealing with situations in the event of high rainfall. The model that is often used in forecasting is the SARIMA (Seasonal Autoregressive Integrated Moving Average) model because it is considered quite simple. However, the SARIMA model has the disadvantage of only being able to capture linear data patterns. In fact, not all data is linear including rainfall. In this research, hybrid SARIMA modeling with LSTM (Long Short Term Memory) artificial neural network is conducted to overcome the problem of non-linear data. Hybrid modeling is expected to improve the accuracy of forecasting results. The final results show that the SARIMA-LSTM hybrid model has better results and accuracy values compared to the SARIMA model.

Item Type:	Thesis (S1)
Additional Information:	ID SINTA Dosen Pembimbing Fitriani Agustina : 5981275 Lukman : 6675529
Uncontrolled Keywords:	SARIMA, LSTM, MAPE SARIMA, LSTM, MAPE
Subjects:	L Education > L Education (General)
Divisions:	Fakultas Pendidikan Matematika dan Ilmu Pengetahuan Alam > Jurusan Pendidikan Matematika > Program Studi Matematika (non kependidikan)
Depositing User:	Annisa Nur Azizah
Date Deposited:	01 Sep 2023 07:34
Last Modified:	01 Sep 2023 07:34
URI:	http://repository.upi.edu/id/eprint/101331

Actions (login required)

View Item