21/3 – Artificial Neural Networks

This module aims to give a solid knowledge of the basic concepts, techniques, and usage of simple artificial neural networks, especially the feedforward dense and convolutional neural networks, which have become standard models used for image categorization and segmentation. We will concentrate on techniques and their application rather than on rigorous mathematical background, and the module relies on programming and hands-on experimenting, so it is accessible to a wider audience. The module serves as an introduction for students with no or limited knowledge of the topic. Since machine learning and specifically deep learning is becoming more and more ubiquitous, anyone interested in it would be warmly welcomed, of course, everyone should take into consideration the hardness of self-assessment exercises. The self-assessment test can be accessed at Entry and Self Assessment Tests -> Files -> Artificial_Neural_Networks_self_assessment.pdf.