SULAIMAN, AAN (2024) PERBANDINGAN ALGORITMA RANDOM FORESTS WITH SPARSE RANDOM PROJECTION (RFDS) DAN RANDOM MULTIMODEL ENSEMBLE UNTUK DETEKSI KONDISI TANAH HUTAN MANGROVE BERDASARKAN EKSTRAKSI FITUR SOIL-ADJUSTED VEGETATION INDEX (Studi Kasus: Kawasan Mangrove Pantai Indah Kapuk). S1 thesis, Universitas Mercu Buana Jakarta.
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (HAL COVER)
01 Cover.pdf Download (450kB) |
|
Text (ABSTRAK)
02 Abstrak.pdf Download (74kB) |
|
Text (BAB I)
03 Bab 1.pdf Restricted to Registered users only Download (184kB) |
|
Text (BAB II)
04 Bab 2.pdf Restricted to Registered users only Download (432kB) |
|
Text (BAB III)
05 Bab 3.pdf Restricted to Registered users only Download (189kB) |
|
Text (BAB IV)
06 Bab 4.pdf Restricted to Registered users only Download (1MB) |
|
Text (BAB V)
07 Bab 5.pdf Restricted to Registered users only Download (73kB) |
|
Text (DAFTAR PUSTAKA)
08 Daftar Pustaka.pdf Restricted to Registered users only Download (148kB) |
|
Text (LAMPIRAN)
09 Lampiran.pdf Restricted to Registered users only Download (984kB) |
Abstract
The mangrove area at Pantai Indah Kapuk has an important role in maintaining the balance of the coastal ecosystem. This research aims to evaluate two methods for detecting mangrove forest soil conditions based on Soil-Adjusted Vegetation Index (SAVI) Characteristic Extraction, namely Random Multimodel Ensemble (RME) and Random Forests with Sparse Random Projection (RFDS). Soil conditions are becoming critical due to human activities and climate change, so accurate technology is needed for monitoring. The research method uses Landsat 8 satellite data via Google Earth Engine for SAVI extraction. Evaluation is carried out by measuring precision, recall and accuracy. The results showed that RFDS had higher precision, recall, and accuracy (0.98, 0.98, and 0.98) than RME (0.96, 0.95, and 0.95). MAE, MSE, RMSE, MAPE, and R2 analysis also shows better performance on RFDS, with MAE of 0.02, MSE 0.03, RMSE 0.19, MAPE 1.3, and R2 0.8. This research provides a significant contribution in understanding the condition of mangrove forest soil in the Pantai Indah Kapuk Mangrove Area, DKI Jakarta. By applying the RFDS method, mangrove soil condition detection technology can be improved accurately and effectively. These findings support the development of technology-based solutions to monitor and protect mangrove ecosystems. Further research could focus on integrating data from other sources and spatial modeling to improve detection accuracy. Thus, the results of this research have the potential to pave the way towards sustainable management and conservation of mangroves in coastal areas. Keywords: Soil Adjusted Vegetation Index, Random Forest, Sparse Random Projection, Google Earth Engine, Landsat 8 Kawasan mangrove di Pantai Indah Kapuk mempunyai peranan penting dalam menjaga keseimbangan ekosistem pesisir. Penelitian ini bertujuan untuk mengevaluasi dua metode deteksi kondisi tanah hutan mangrove berdasarkan Ekstraksi Ciri Soil-Adjusted Vegetation Index (SAVI), yaitu Random Multimodel Ensemble (RME) dan Random Forests with Sparse Random Projection (RFDS). Kondisi tanah menjadi kritis akibat aktivitas manusia dan perubahan iklim, sehingga diperlukan teknologi yang akurat untuk pemantauannya. Metode penelitian menggunakan data satelit Landsat 8 melalui Google Earth Engine untuk ekstraksi SAVI. Evaluasi dilakukan dengan mengukur presisi, recall dan akurasi. Hasil penelitian menunjukkan bahwa RFDS memiliki presisi, recall, dan akurasi yang lebih tinggi (0.98, 0.98, dan 0.98) dibandingkan RME (0.96, 0.95, dan 0.95). Analisis MAE, MSE, RMSE, MAPE, dan R2 juga menunjukkan kinerja yang lebih baik pada RFDS, dengan MAE sebesar 0.02, MSE 0.03, RMSE 0.19, MAPE 1.3, dan R2 0.8. Penelitian ini memberikan kontribusi yang signifikan dalam memahami kondisi tanah hutan mangrove di Kawasan Mangrove Pantai Indah Kapuk, DKI Jakarta. Dengan menerapkan metode RFDS, teknologi deteksi kondisi tanah mangrove dapat ditingkatkan secara akurat dan efektif. Temuan ini mendukung pengembangan solusi berbasis teknologi untuk memantau dan melindungi ekosistem mangrove. Penelitian lebih lanjut dapat fokus pada pengintegrasian data dari sumber lain dan pemodelan spasial untuk meningkatkan akurasi deteksi. Dengan demikian, hasil penelitian ini berpotensi membuka jalan menuju pengelolaan berkelanjutan dan konservasi mangrove di wilayah pesisir. Kata Kunci : Soil Adjusted Vegetation Index, Random Forest, Sparse Random Projection, Google Earth Engine, Landsat 8
Actions (login required)
 |
View Item |
