Speaker: Prof. Robert Jenssen (Director of Visual Intelligence Research Centre)
Title: Deep learning in image analysis, graphs, and a new measure of statistical dependency between graphs
Abstract: This talk will provide examples of deep learning in image analysis and how the concept of convolutional neural networks can be extended to graphs. The talk then shows an example of graph neural network learning where dependency measures between graphs are needed, for then to present a new approach based on information theoretic learning developed in the UiT Machine Learning Group and the Visual Intelligence research centre.