Prediksi Nilai Tanah Menggunakan Geographically Weighted Extreme Learning Machine

Wicaksono, Eko Arif (2024) Prediksi Nilai Tanah Menggunakan Geographically Weighted Extreme Learning Machine. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022221007-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (2MB) | Request a copy

Abstract

Dengan kemajuan teknologi, ada ekspektasi bahwa teknologi dapat menyederhanakan dan mempercepat proses penilaian tanah, terutama dengan memanfaatkan metode pembelajaran mesin untuk memprediksi nilai tanah. Penelitian ini mengeksplorasi pendekatan Geographically Weighted Extreme Learning Machine (GWELM) untuk memprediksi nilai tanah. GWELM menggabungkan metode Extreme Learning Machine (ELM) dengan pembobotan geografis dari Geographically Weighted Regression (GWR). Hasil evaluasi menunjukkan bahwa model GWELM memiliki akurasi prediksi yang lebih baik dibandingkan dengan model Support Vector Machine (SVM), Convolutional Neural Network (CNN), dan ELM tanpa pembobotan geografis. Dari hasil evaluasi, model GWELM menunjukkan penurunan Mean Absolute Error (MAE) sebesar 32,26%, penurunan Mean Absolute Percentage Error (MAPE) sebesar 23,42%, dan peningkatan koefisien determinasi (R2) sebesar 895,92% dibandingkan dengan model SVM. Dengan CNN, GWELM menunjukkan penurunan MAE sebesar 13,82% dan MAPE sebesar 6,80%, serta peningkatan R² sebesar 52,85%. Dibandingkan dengan ELM, GWELM juga menunjukkan penurunan MAE sebesar 2,89% dan MAPE sebesar 1,49%, dengan peningkatan R² sebesar 11,48%. Dengan kombinasi metode ELM dan GWR, prediksi nilai tanah menjadi lebih akurat, meskipun memerlukan biaya komputasi yang lebih tinggi dibandingkan dengan model ELM tanpa pembobotan geografis. Penelitian ini menggunakan data geospasial yang disimpan dalam format Shapefile (SHP), salah satu format data geospasial yang paling umum digunakan dalam Sistem Informasi Geografis (GIS). Untuk memproses dan mengekstrak informasi yang diperlukan dari data ini, diperlukan alat khusus seperti ArcGIS, yang memungkinkan ekstraksi koordinat dan fitur lainnya. Penelitian ini menunjukkan bagaimana pendekatan GWELM dapat meningkatkan proses prediksi nilai tanah dengan mempertimbangkan aspek geografis dari data tersebut.
==================================================================================================================================
With the advancement of technology, there is an expectation that technology can simplify and accelerate the land valuation process, especially by utilizing machine learning methods to predict land value. This research explores the Geographically Weighted Extreme Learning Machine (GWELM) approach to predicting land values. GWELM combines the Extreme Learning Machine (ELM) method with geographical weighting from Geographically Weighted Regression (GWR). The evaluation results show that the GWELM model has better prediction accuracy compared to the Support Vector Machine (SVM), Convolutional Neural Network (CNN), and ELM models without geographic weighting. From the evaluation results, the GWELM model shows a decrease in Mean Absolute Error (MAE) of 32.26%, a decrease in Mean Absolute Percentage Error (MAPE) of 23.42%, and an increase in the coefficient of determination (R2) of 895.92% compared to the SVM model. With CNN, GWELM showed a decrease in MAE by 13.82% and MAPE by 6.80%, as well as an increase in R² by 52.85%. When compared to ELM, GWELM also showed a decrease in MAE by 2.89% and MAPE by 1.49%, with an increase in R² by 11.48%. With the combination of ELM and GWR methods, land value prediction becomes more accurate, although it requires higher computational costs compared to the ELM model without geographic weighting. This research uses geospatial data stored in Shapefile (SHP) format, one of the most commonly used geospatial data formats in Geographic Information Systems (GIS). To process and extract the necessary information from this data, specialized tools such as ArcGIS are required, which allow the extraction of coordinates and other features. This research shows how the GWELM approach can improve the land value prediction process by considering the geographical aspects of the data.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Prediksi Nilai Tanah, Prediction Land Value, GWELM, ELM, GWR.
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Eko Arif Wicaksono
Date Deposited:	26 Jul 2024 07:14
Last Modified:	26 Jul 2024 07:14
URI:	http://repository.its.ac.id/id/eprint/109124

Actions (login required)

View Item