Prediksi peluang kejadian gempa bumi di Aceh dengan Poisson Hidden Markov Model = Predicting earthquake probabilites in Aceh using the Poisson Hidden Markov Model

Leodinata, Elisse (2024) Prediksi peluang kejadian gempa bumi di Aceh dengan Poisson Hidden Markov Model = Predicting earthquake probabilites in Aceh using the Poisson Hidden Markov Model. Bachelor thesis, Universitas Pelita Harapan.

	Text (Title) Title.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (244kB)
	Text (Abstract) Abstract.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (577kB)
	Text (ToC) ToC.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (583kB)
	Text (Chapter1) Chapter1.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (581kB)
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (707kB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (673kB)
	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 (557kB)
	Text (Bibliography) Bibliography.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (559kB)
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (6MB)

Abstract

Penelitian ini menangani tantangan prediksi gempa bumi di wilayah Aceh melalui penerapan Poisson hidden Markov model (PHMM). Tujuan utama melibatkan penentuan parameter optimal PHMM melalui estimasi Bayesian posterior (BPM), pembuatan model prediktif untuk probabilitas gempa bumi di Aceh, dan evaluasi kinerja model tersebut pada berbagai periode waktu. Penelitian ini menggunakan metode Gibbs sampler untuk mengestimasi parameter PHMM, mengeksplorasi model dengan dua, tiga, dan empat hidden states. Bertentangan dengan harapan, pembagian data observasi ke dalam periode waktu yang berbeda tidak memberikan hasil yang signifikan dalam hasil prediksi. Namun, metode Gibbs sampler terbukti efektif dalam estimasi parameter. Secara mencolok, model PHMM menunjukkan prediksi yang akurat untuk frekuensi gempa bumi berintensitas rendah tetapi cenderung meremehkan frekuensi gempa bumi berintensitas tinggi. Signifikansi dari penelitian ini terletak pada kontribusinya terhadap peningkatan metodologi prediksi gempa bumi di Aceh. Temuan ini bertujuan untuk meningkatkan akurasi ramalan, khususnya untuk peristiwa seismik dengan intensitas lebih tinggi. Dengan mengevaluasi dan menyempurnakan model PHMM, penelitian ini mendukung upaya berkelanjutan untuk meningkatkan kesiapsiagaan dan strategi mitigasi gempa bumi di daerah rawan gempa. / This study addresses the challenge of earthquake prediction in the Aceh region through the application of the Poisson Hidden Markov Model (PHMM). The primary objectives include determining the optimal PHMM parameters through Bayesian Parameter Estimation (BPM), constructing a predictive model for earthquake probabilities in Aceh, and evaluating the model’s performance across different time frames. The study employs the Gibbs sampler method to estimate PHMM parameters, exploring models with two, three, and four hidden states. Contrary to expectations, the division of observation data into different time frame periods yields non-significant results in forecasting. However, the Gibbs sampler method proves effective in parameter estimation. Notably, PHMM models demonstrate accurate predictions for low-intensity earthquake frequencies but exhibit underestimation for high-intensity earthquake frequencies. The significance of this research lies in its contribution to the enhancement of earthquake prediction methodologies in Aceh. The findings aim to improve forecasting accuracy, particularly for seismic events with higher intensities. By evaluating and refining PHMM models, this study supports ongoing efforts to enhance earthquake preparedness and mitigation strategies in earthquake-prone regions.

Actions (login required)

View Item