**Information for enrolled/prospective students**: all information is on SKOS (please self-enroll into Deep Learning 2021), materials are on GitHub, lectures will be streamed and recorded in Teams. **Enrollment code** is in SKOS.
Neural Networks and other Deep Learning techniques allow creation of programs
that are learned rather than written. This means that instead of implementing
a concrete algorithm, the program applies patterns that are automatically
found in the data. In example, a translation system such as the Google
Translate can be created by applying a deep neural network to a large corpus
of translated documents.
The lecture is a continuation of the Machine Learning course, focusing on Deep Learning techniques. We will speak about the current state of the art techqniques for image recognition and language processing such as convolutional and recurrent neural networks with the attention mechanism. We will also speak about data generation techniques, such as autoencoders and generative adversarial networks and we wil see a bit of deep reinforcement learning.
The lecture will be accompanied by computer exercises:
* Key topics will be illustrated by small exercises.
* We wil introduce the PyTorch deep learning framework.
* A larger project will give you the opportunity to tackle a real world machine learning problem.
