We are using Artificial Neural Network (ANN) analytical methods to identify emergent pest problems and create risk indices for specific species.

ANNs are computational tools that have novel application to problems in invasion biology.

Self organising maps have been used to analyse a database comprising the world distribution of 844 insect pest species to rank their establishment potential in New Zealand based on the strength of their association with geographic pest assemblages. Back-propagation neural networks have been used to further inform the risk analysis process by identifying variables that best explain current species distribution and use that information to predict presence in areas where the species is not normally found.

To date, two predictive models have been developed. We are combining them with statistical and computational methods to model potential invasive species establishment and spread throughout New Zealand. 

Back to Intelligent Systems for Biosecurity homepage