Zahro, Eling Diana (2023) Analisis Perubahan Tutupan Lahan Tahun 2019 dan 2021 dengan Metode Support Vector Machine (SVM) dan Classification and Regression Tree (CART) (Studi Kasus : Kabupaten Kubu Raya, Provinsi Kalimantan Barat). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 03311940000034_Undegraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
03311940000034_Undegraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (4MB) | Request a copy |
Abstract
Perubahan tutupan lahan di Kabupaten Kubu Raya selama pandemi COVID-19 khususnya pada tahun 2019 sampai 2021pada saat duniamengalami lockdown dimana segala aktivitas manusia dibatasi meliputi mobilitas, aktivitas sosial, dan aktivitas manusia yang berhubungan secara langsung. Dengan adanyaadanya pandemiinitentu sajamempengaruhi pembangunan-pembangunan yang mengakibatkan peran manusia secara langsung. Pada penelitian ini klasifikasi penutup lahan di Kabupaten Kubu Raya dilakukan dengan mengambil metode machine learning terbaik dari penelitian terdahulu yang telah dilakukan yaitu dengan algoritma Support Vector Machine (SVM) dan Decision tree CART. Citra satelit yang digunakan adalah citra SPOT 6 dan 7 tahun 2019 dan 2021. Perbandingan splitting ratio yang digunakan dalam pembuatan desain training point berturut-turut perbandingan data training dan data testing adalah 60:40; 70:30; 75:25; 80:20; dan 90:10. Hasil perhitungan menunjukkan bahwa metode Support Vector Machine dan Classification and Regression Tree, didapatkan lima kelas tutupan lahan yaitu area terbangun, badan air, lahan kosong, pertanian, dan vegetasi untuk tahun 2019 dan 2021 dengan perbandingan komposisi spliting ratio sebesar 75 % training data dan 25 % testing data. Berdasarkan uji akurasi yang dilakukan secara kuantitatif menggunakan confusion matrix, didapatkan bahwa metode Support Vector Machine memiliki akurasi yang lebih tinggi dan memenuhi ketentuan USGS dibandingkan dengan metode Classification and Regression Tree dengan nilai overall accuracy dan kappa accuracy baik pada tahun 2019 maupun 2021 secara berturut turut 0,9615 ; 0,9487 dan 0,9221 ; 0,8955. Hasil pemodelan tutupan lahan kedua metode menunjukkan bahwa metode SVM lebih akurat dibandingkan dengan metode CART.
=================================================================================================================================
Changes in land cover in Kubu Raya Regency during the COVID-19 pandemic, especially in 2019 to 2021 when the world is experiencing a lockdown where all human activities are limited including mobility, social activities, and human activities that are directly related. With the existence of this pandemic, it affects developments that result in a direct human role. In this study, land cover classification in Kubu Raya Regency was carried out by taking the best machine learning method from previous studies that had been carried out, the Support Vector Machine (SVM) algorithm and CART Decision Tree. The satellite images used are SPOT 6 and 7 images for 2019 and 2021. The splitting ratio used in making the training point design respectively the comparison of training data and testing data is 60:40; 70:30; 75:25; 80:20; and 90:10. The calculation results show that with the Support Vector Machine and Classification and Regression Tree methods, five land cover classes are obtained, known as builtup areas, water bodies, vacant land, agriculture, and vegetation for 2019 and 2021 with a splitting ratio composition of 75% training data and 25 % testing data. Based on the accuracy test carried out quantitatively using the confusion matrix, it was found that the Support Vector Machine method has higher accuracy and meets the USGS requirements compared to the Classification and Regression Tree method with overall accuracy and kappa accuracy values in both 2019 and 2021 respectively 0 ,9615 ; 0.9487 and 0.9221 ; 0.8955. The results of the land cover modeling of the two methods show that the SVM method is more accurate than the CART method.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSG 333.731 3 Zah a-1 |
| Uncontrolled Keywords: | SVM, Decision tree CART, Machine learning |
| Subjects: | H Social Sciences > HT Communities. Classes. Races > HT133 City and Towns. Land use,urban Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. |
| Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis |
| Depositing User: | Eling Diana Zahro |
| Date Deposited: | 11 Sep 2023 03:37 |
| Last Modified: | 18 Dec 2023 05:12 |
| URI: | http://repository.its.ac.id/id/eprint/101532 |
Actions (login required)
 |
View Item |