Skip to main content

ORIGINAL RESEARCH article

Front. Environ. Sci.
Sec. Land Use Dynamics
Volume 12 - 2024 | doi: 10.3389/fenvs.2024.1431645

A Performance Evaluation of Random Forest, Artificial Neural Network, and Support Vector Machine Learning Algorithms to Predict Spatio-temporal Land Use-Land Cover Dynamics: A Case from Lusaka and Colombo

Provisionally accepted
Bwalya Mutale Bwalya Mutale 1NEEL CHAMINDA WITHANAGE NEEL CHAMINDA WITHANAGE 2,3Dr. Prabuddh K. Mishra Dr. Prabuddh K. Mishra 4*Jingwei Shen Jingwei Shen 2Kamal Abdelrahman Kamal Abdelrahman 5Mohammed S. Fnais Mohammed S. Fnais 5
  • 1 Liaoning Technical University, Fuxin, Liaoning Province, China
  • 2 Southwest University, Chongqing, Chongqing Municipality, China
  • 3 University of Ruhuna, Matara, Sri Lanka
  • 4 Shivaji College, University of Delhi, New Delhi, India
  • 5 King Saud University, Riyadh, Riyadh, Saudi Arabia

The final, formatted version of the article will be published soon.

    Reliable information is crucial for sustainable urban planning. Advances in computer technology and geoinformatics tools enable accurate identification of land use and land cover (LULC) in both spatial and temporal dimensions. Precise information is essential for better decision-making, making it crucial to assess the performance of classification algorithms in detecting LULC changes. Although widely studied in many countries, research on machine learning algorithms for LULC evaluation is limited in Zambia and Sri Lanka. We aimed to assess the reliability and performance of Support Vector Machine (SVM), Random Forest (RF), and Artificial Neural Network (ANN) algorithms in detecting LULC changes in Lusaka and Colombo City from 1995 to 2023, using Landsat Thematic Mapper (TM) and Operational Land Imager (OLI). The results show that RF and ANN models performed best, both achieving a Mean Overall Accuracy (MOA) of 96% for Colombo, and 96% and 94% for Lusaka, respectively. The SVM model had an Overall Accuracy (OA) of 77% to 94% between 1995 and 2023. The RF algorithm outperformed ANN and SVM models, achieving higher overall accuracy (OA) and kappa coefficients (0.92 to 0.97) across both study areas. Significant LULC changes were observed: Lusaka saw a 60.4% increase in vegetation (11,990 hectares) from 1995 to 2005, primarily converting bare land to vegetation. From 2005 to 2023, built-up areas grew by 71% (25,110 hectares), with a net vegetation gain of over 11,000 hectares (53.4%) from 1995 to 2023. In Colombo, built-up areas increased by 81.5% (1,779 hectares) while vegetation decreased by 1,519 hectares (62.3%). LULC simulations predict further expansions of built-up areas by 160 hectares in Lusaka and 337 hectares in Colombo from 2023 to 2035. The RF algorithm outperformed ANN and SVM, and the resulting land cover maps will be invaluable for urban planning and policy development in both countries.

    Keywords: artificial neural network, Colombo, Land use/land cover, Lusaka, random forest, Support Vector Machine Bibliography1, Justified, Indent: Left: 0"

    Received: 12 May 2024; Accepted: 27 Aug 2024.

    Copyright: © 2024 Mutale, WITHANAGE, Mishra, Shen, Abdelrahman and Fnais. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence: Dr. Prabuddh K. Mishra, Shivaji College, University of Delhi, New Delhi, India

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.