Course introduction
Machine learning has become a part in our everyday lifes, from simple product recommendations to personal electronic assistants to self-driving cars. Especially Deep Learning has gained a lot of interest in the media and has demonstrated impressive results. This intensive course will introduce the student to the exciting world of deep learning. We will learn about the theoretical background and concepts driving deep learning and highlight and discuss the most noteworthy applications of deep learning but also their limitations. Furthermore, all content will immediately put into practice by suitable exercises and programming tasks.
In relation to the competence profile of the degree it is the explicit focus of the course to:
Give knowledge and understanding of a collection of specialized models and methods developed within Computer Science based on research on highest international level, as well as of models and methods aimed at applications in other subject areas.
Give skills to describe, analyze and solve computational problems by using the methods learnt, to analyze pros and cons of different methods in Computer Science, as well as to develop new variants of the methods learnt where the problem at hands requires this.
Give the competence to plan and execute scientific projects on a high technical level.
Expected learning outcome
The learning objectives of the course is that the student demonstrates the ability to:
Describe the principles of deep neural networks in a scientific and precise language and notation
Analyze the various types of neural networks, the different layers and their interplay
Discuss the feasibility of deep learning approaches to concrete problems
Describe the theoretical mathematical foundations of the field
Implement and apply deep learning frameworks for solving concrete problems
Utilize state-of-the-art deep learning frameworks for implementing deep neural networks
Content
The following main topics are contained in the course:
feedforward neural networks
recurrent neural networks
convolutional neural networks
backpropagation algorithm
regularization
