Kursusinformation

Artificial Intelligence (AI, in particular Machine Learning, Big Data, Neural Networks) is generally considered a key technology for the near future. Currently, AI methods are used e.g. in games (e.g. navigation, conversation, strategy, adaptation of virtual characters), human computer interaction (recognition of faces, voices, speech), robotics (navigation), music information retrieval  (detecting pitch, harmony, rhythm, instrument), and consumer behavior analysis in social networks. This course will cover in particular pathfinding algorithms, decision making, steering behavior, basic graph theory, classification, (convolutional) neural networks, and reinforcement learning. The course includes practical programming exercises in various areas, e.g. game AI and image recognition. 

The course includes the following topics: finite state automata, path finding algorithms, search strategies, graph theory, decision trees, agents, introduction to classification, convolutional neural networks, recurrent neural networks, reinforcement learning. A special emphasis is put on the programming experience in this course.

This course builds on previous programming (Introduction to Programming, B.Sc. Medialogy  1^st, Computer Graphics Programming, B.Sc. Medialogy  5^th), mathematics (Mathematics for Multimedia Applications, B.Sc. Medialogy 2^nd) and statistics classes (Design and Analysis of Experiments, B.Sc. Medialogy 4^th), and complements Theory and Practice of Game Design and Development,  Technologies for Web and Social Media (also Medialogy B.Sc. 6^th Semester). The course provides an introduction to the course Machine Learning for Multimedia Applications (SMC/MED M.Sc. 1^st Semester) and a preparation e.g. for music information retrieval taught in the course Sound and Music Signal Analysis (SMC M.Sc. 2^nd Semester) and for automatic gesture recognition that is relevant to the courses Multimedia Programming and  Embodied Interaction  (Medialogy M.Sc. 2^nd Semester).