Detecting and interpreting patterns within regional pest species assemblages using self-organizing maps and other clustering methods.
2015
Worner , Sue | Eschen , Rene | Kenis, Marc | Paini, Dean | Saikkonen, Kari | Suiter, Karl | Singh, Sunil K. | Vanninen, Irene | Watts, Mike | Venette, Robert C.
This chapter highlights quantitative methods designed to identify and rank exotic species with potential risk to cause economic and/or environmental harm if they establish in a new area. Until now, pest risk assessments have tended to be qualitative and reactive instead of quantitative and proactive. Here, a computational-intelligence technique called a self-organizing map (SOM) is described that can be used to analyse regional profiles or assemblages of pest species to determine their potential for establishment in new regions. In addition to the SOM, two other useful clustering or classification algorithms, k-means and hierarchical analysis, are also demonstrated to provide a quantitative framework to the risk assessment process. The examples described for each method illustrate how a pest risk analyst can identify, from a large list of potential hazards, which species present the most risk to target areas. Furthermore, examples are given of how such analyses may indicate donor and recipient regions for pest invasion and can highlight previously unknown or ignored threats for further investigation. Finally, cautions are provided and limitations of SOMs and other clustering methods applied to the area of pest risk assessment are discussed.
Afficher plus [+] Moins [-]Informations bibliographiques
Cette notice bibliographique a été fournie par University of the South Pacific
Découvrez la collection de ce fournisseur de données dans AGRIS
