Dharmawan, Bintang (2024) Analisis Prediksi Penurunan Lahan Gambut Menggunakan Algoritma Extreme Gradient Boost (Studi Kasus: Kalimantan Selatan). Other thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
5016201120_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (6MB) | Request a copy |
Abstract
Penelitian ini mengeksplorasi penerapan teknologi algoritma extreme gradient boost (XGBoost) untuk memprediksi penurunan permukaan lahan gambut di Kabupaten Banjar, Banjar Baru, dan Tanah Laut, Kalimantan Selatan, Indonesia. Lahan gambut memiliki peran penting dalam penyimpanan karbon, pelestarian keanekaragaman hayati, dan pengaturan air, tetapi terancam oleh subsidensi, yang sering kali diperparah oleh aktivitas manusia dan kebakaran hutan. Penelitian ini bertujuan untuk membuat model peramalan dengan algoritma XGBoost dengan menggunakan data deret waktu INSAR, yang berkontribusi pada pemantauan penurunan permukaan lahan gambut secara real-time dan tepat. Tujuan penelitian ini termasuk memahami bagaimana hasil prediksi subsidensi lahan gambut dan akurasinya ketika divalidasi menggunakan data INSAR. Hasil penelitian menunjukkan bahwa akurasi terbaik dari model yang dievaluasi dengan menggunakan RMSE dan MAE adalah 5,2155 mm atau 0,52 cm dan 3,787 mm atau 0,37 cm menunjukkan hasil yang cukup baik. Hasil penelitian ini memberikan wawasan baru mengenai model subsidensi lahan gambut di Kabupaten Banjar, Banjar Baru, dan Tanah Laut, Kalimantan Selatan, serta potensi penelitian serupa di seluruh dunia.
=================================================================================================================================
This research explores the application of extreme gradient boost (XGBoost) algorithm technology to predict peatland subsidence in Banjar, Banjar Baru, & Tanah Laut Regency in South Kalimantan, Indonesia. Peatlands play a crucial role in carbon storage, biodiversity preservation, and water regulation, but are threatened by subsidence, often exacerbated by human activities and forest fires. The research aims to create forecasting models with XGBoost algorithm by using INSAR time-series data, contributing to real-time and precise monitoring of peatland subsidence. The research objectives include understanding how the prediction result of peatland subsidence and its accuracy when validate using INSAR data. The findings show the best accuracy of the model evaluate using both RMSE and MAE was 5.2155 mm or 0.52 cm and 3.787 mm or 0.37 cm showing quite good result. It offers new insights into the model of peatland subsidence in Banjar, Banjar Baru, & Tanah Laut Regency in South Kalimantan and potentially similar research worldwide.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | INSAR, XGBoost, peatland subsidence, penurunan lahan gambut |
Subjects: | G Geography. Anthropology. Recreation > G Geography (General) G Geography. Anthropology. Recreation > G Geography (General) > G70.217 Geospatial data G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing G Geography. Anthropology. Recreation > GE Environmental Sciences |
Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis |
Depositing User: | Bintang Dharmawan |
Date Deposited: | 29 Jul 2024 08:33 |
Last Modified: | 29 Jul 2024 08:33 |
URI: | http://repository.its.ac.id/id/eprint/109418 |
Actions (login required)
![]() |
View Item |