University of Tasmania
Browse

A validated ensemble method for multinomial land-cover classification

Download (876.22 kB)
Version 2 2025-01-17, 05:27
Version 1 2023-05-20, 10:50
journal contribution
posted on 2025-01-17, 05:27 authored by VL Diengdoh, Stefania OndeiStefania Ondei, Mark HuntMark Hunt, Barry BrookBarry Brook
Land-cover data provides valuable information for landscape management and can be generated using machine learning algorithms. Ensemble models or model averaging can overcome difficulties in selecting an adequate algorithm and improve model predictions, but its use is limited among ecologists. The objective of this study is to highlight the benefits and limitations of weighted and unweighted majority voting ensemble models for land-cover classification and to enable easy and wider implementation of the method by providing an R-script (for use in the R software). Using a case study of three mixed-use landscapes from southern Australia (Tasmania), land cover was classified into six classes using Landsat 8 imagery and ancillary data, and support vector machine, random forest, k-nearest neighbour and naïve Bayesian as base algorithms. The predicted classifications of the base algorithms were then averaged using both an unweighted and weighted (using the true skill statistic) majority voting ensemble algorithm. Cross-validation results showed the base algorithms achieved similar accuracy making algorithm selection difficult. The base algorithms achieved high and similar predictive accuracy when the classified land-cover and training data belong to the same geographic region but lower and different predictive accuracy when the classified land-cover and training data belong to different geographic regions. The weighted and unweighted ensemble achieved similar overall accuracy, equivalent to the best performing base algorithm. We conclude that the majority voting ensemble can be adopted to overcome difficulties in model selection during land-cover classification.

History

Publication title

Ecological Informatics

Volume

56

Article number

101065

Number

101065

Pagination

1-10

ISSN

1574-9541

Department/School

Biological Sciences

Publisher

Elsevier Science Bv

Publication status

  • Published

Place of publication

Netherlands

Rights statement

© 2020 Elsevier B.V. All rights reserved. This is the author accepted version of the published article https://doi.org/10.1016/j.ecoinf.2020.101065 made accessible in this repository under the terms of the publisher's author self-archiving policy.

Socio-economic Objectives

180403 Assessment and management of Antarctic and Southern Ocean ecosystems, 189999 Other environmental management not elsewhere classified

UN Sustainable Development Goals

15 Life on Land