Líffræðifélag Íslands - biologia.is
Líffræðiráðstefnan 2021

Erindi/veggspjald / Talk/poster E23

Wildlife ecological modeling in Iceland: How scientists can leverage Machine Learning

Höfundar / Authors: Thomas Y. Chen (1)

Starfsvettvangur / Affiliations: 1. Academy for Mathematics, Science, and Engineering

Kynnir / Presenter: Thomas Y. Chen

Iceland is a biodiverse country with various species of mammals, birds, and fish. To model the ecological systems and relationships that exist and in order to develop mechanisms for conservation and preservation, machine learning and artificial intelligence-based biodiversity informatics can be a key tool. Machine learning consists of training computer programs to gain insights from big data that can be applied via inference to unseen data. Techniques applicable to Icelandic ecology include regression, clustering, decision trees, random forests, neural networks, etc. Applications include deep learning-based computer vision for wildlife identification in imagery, which helps scientists to understand population and species-level trends in a noninvasive manner. In this talk, we overview machine learning / AI techniques that can advance biodiversity research in Iceland, as climate change continues to threaten ecosystems.