Machine perception of natural signals has improved a lot in the recent years thanks to deep learning (DL). Improved image recognition with DL will make self-driving cars possible and is leading to more accurate image-based medical diagnosis. Improved speech recognition and natural language processing with DL will lead to many new intelligent applications within health-care and IT. Pattern recognition with DL in large datasets will give new tools for drug discovery, condition monitoring and many other data-driven applications.
The purpose of this course is to give the student a detailed understanding of the deep artificial neural network models, their training, computational frameworks for deployment on fast graphical processing units, their limitations and how to formulate learning in a diverse range of settings. These settings include classification, regression, sequences and other types of structured input and outputs and for reasoning in complex environments.
