Link Prediction for Ecological Networks
In computer science, link prediction methods have been developed to understand connections between entities in a network (e.g., social networks). In particular, recommendation algorithms such as latent factor models analyze links in bipartite networks and have been applied to many industrial applications (e.g., Netflix and Amazon), showing significant benefits by predicting new connections. However, these technologies have not yet been applied to ecological networks. In this proposal, we propose to incorporate recommendation algorithms into ecological models in order to understand interactions between species. Because ecological data is fundamentally different from other data in computer science, we address several challenges in analyzing ecological networks and propose various methods that handle each of the challenges.
Major Advisor: Rebecca Hutchinson
Committee: Thomas Dietterich
Committee: Xiaoli Fern
Committee: Xiao Fu
GCR: Brett Tyler
Tuesday, June 9, 2020 at 11:00am to 1:00pm
Virtual EventDakota Nelson
No recent activity