Penerapan algoritma kalman filter dalam perbaikan hasil prediksi cuaca dengan model ARIMA

Gabriela, Natasha (2021) Penerapan algoritma kalman filter dalam perbaikan hasil prediksi cuaca dengan model ARIMA. Bachelor thesis, Universitas Pelita Harapan.

	Text (Title) Title - 2021-02-22T113229.784.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (345kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (347kB) | Preview
Preview	Text (Chapter1) Chapter1.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (349kB) | Preview
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (562kB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (687kB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (332kB)
Preview	Text (Bibliography) Bibliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (335kB) | Preview
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB)

Abstract

Mata pencaharian masyarakat Bandar Lampung didominasi oleh sektor pertanian, kehutanan, dan perikanan sebesar 47.97%. Informasi tentang cuaca yang akan terjadi menjadi sangat penting bagi penduduk di Kota Bandar Lampung. Pada penelitian ini, dilakukan prediksi cuaca menggunakan unsur-unsur cuaca yaitu suhu dan kelembaban. Prediksi cuaca dimasa yang akan datang dapat dilakukan apabila terdapat data suhu dan kelembaban di masa tersebut. Suhu dan kelembaban di masa depan diramalkan terlebih dahulu dengan menggunakan model deret waktu, ARIMA dan ARIMA-KF. ARIMA-KF adalah model ARIMA yang telah dikenakan algoritma Kalman filter. Kedua model dibandingkan, model dengan indikator akurasi terkecil terpilih sebagai model yang hasil peramalannya digunakan untuk memprediksi cuaca. Prediksi cuaca dilakukan dengan menggunakan model regresi logistik biner yang sebelumnya telah divalidasi dengan metode k-fold cross validation. Dibentuk dua model regresi logistik, biasa dan polinomial dimana model polinomial terdiri dari suku kuadratik dan interaksi dari variabel independen. Analisis hasil prediksi cuaca menunjukkan bahwa hasil peramalan deret waktu terpilih ARIMA-KF tidak berbeda secara signifikan dengan data aktual serta akurasi hasil prediksi dengan hasil peramalan deret waktu tidak akan melampaui nilai akurasi hasil prediksi dengan data aktual. / The structure of the working population according to employment in Bandar Lampung is dominated by agriculture, forestry and fisheries amounting to 47.97%. It means that residents need sufficient information about the weather that occurs every day. In this study, a weather prediction is carried out using weather elements, temperature and humidity. Future weather predictions can be done if there is temperature and humidity data at that time. Future temperature and humidity are predicted in advance using the time series models, ARIMA and ARIMA-KF. ARIMA-KF is the ARIMA model which has been subjected to the Kalman filter algorithm. Comparing the two models, model with the smallest accuracy indicator is selected as the model whose forecasting results are used to predict the weather. Weather prediction is carried out using a binary logistic regression model that has been previously validated using the k-fold cross validation method. Simple logistic and polynomial logistic regression models were formed, where the polynomial model consisted of quadratic terms and the interaction of the independent variables. The analysis of the weather prediction results shows that the ARIMA-KF time series forecasting results are not significantly different from the actual data and the accuracy of the prediction results with the time series forecasting results will not exceed the accuracy of the predicted results with the actual data.

Actions (login required)

View Item