This project will will use machine learning methods to predict the geographical distribution of wildlife species. Ground truth data for species distribution can be obtained from a citizen science portal such as the National Biodioversity Network. Data to characterise and differentiate between different locations will be obtained from various sources that record environmental features such as climate, land cover, elevation, soil types and population density. The project will experiment with classifiers that use different combinations of features to find the classification approaches that are most effective at predicting the occurrence of particular wildlife species.