Predicting Potential Geographical Distribution of Tobacco Blue Mold in China
DOI:
https://doi.org/10.3329/bjb.v51i40.63830Keywords:
Ecological niche model, Alien species, Support Vector Machine (SVM), GIS, Cross validationAbstract
Tobacco blue mold is a serious tobacco disease, which has high risk of invasion in many areas, therefore, it is of particular importance to strengthen the quarantine work and to make a good prediction. Since each species has its own particular relatively stable ecological niche, one can predict the potential geographical distribution of tobacco blue mold in target areas like China. Based on the known distribution data of tobacco blue mold, 20 environmental factors were selected and combined support vector machine (SVM) with ecological niche model was used to predict the potential spatial distribution of tobacco blue mold and to obtain the optimal combination of SVM parameters through the brute force search algorithm. Search algorithm overcame the shortage of empirical method and improved the accuracy of prediction. The method was proved to be effective and feasible by cross validation which provided a robust theoretical basis for preventing the intrusion and spread of tobacco blue mold.
Bangladesh J. Bot. 51(4): 883-888, 2022 (December) Special
Downloads
33
55
Downloads
Published
How to Cite
Issue
Section
License
© Bangladesh Botanical Society
Authors are required to transfer the copyright of their articles to the journal. The Declaration form is available here http://www.bdbotsociety.org/journal/journal_pdf/declaration_form.pdf