= Deep Neural Networks =

Deep Neural Networks is an optional course in the module "Machine Learning - Theory and Applications" and is worth 3 LP (3 ECTS credits).

In the general case, it is not possible to take the Deep Neural Networks course as a standalone course. There are possible exceptions to this (e.g. it complements another ML or related course you are taking in parallel). In that case, a special request needs to be made. 

=== General Information (preliminary) ===

||'''Lecture period'''|| from 09 May 2018 to 27 Jun 2018 ||
||'''Lecture'''|| Wednesdays 08:15-10:00 in room H 2032 ||
||'''Language'''|| English ||
||'''Trainer'''||Grégoire Montavon||
||'''Contact''' || gregoire.montavon@tu-berlin.de ||
|| '''ISIS''' || https://isis.tu-berlin.de/course/view.php?id=13335 ||

=== Frequently Asked Questions (FAQ) ===

 * When does the course start? On Wednesday 9 May 2018
 * How to register for the course? Pre-registration is not needed.

=== Main Topics ===

 * DNN basics
 * Model complexity and model selection
 * Optimization of DNNs
 * DNNs for images
 * DNNs for time series
 * Explaining DNN decisions
 * DNNs beyond classification