Mittwoch, 27. Juni 2018

[HIForum] [Kolloquium] EINLADUNG / INVITATION - Informatisches Kolloquium 02.07.2018, 17:15, B201, Informatikum/Stellingen

Dies ist eine Einladung zum Informatischen Kolloquium am Montag, den 02. Juli 2018, um 17:15 auf dem Universitätsgelände "Informatikum/Stellingen", im Konrad-Zuse-Hörsaal B-201. Der Vortrag mit dem Titel "Evolution of Convolutional Highway Networks" wird von Herrn Prof. Dr. Oliver Kramer von der Universität Oldenburg gehalten (Vortragssprache: Deutsch/Englisch).

 

This is an invitation to the next informatics colloquium on Monday, 2 July 2018, at 17:15, on university campus Stellingen, lecture hall B-201. The talk is entitled "Evolution of Convolutional Highway Networks" and will be given in English or German (according to preference) by Prof. Dr. Oliver Kramer.

 

Nähere Informationen zu allen in diesem Semester geplanten Vorträgen finden Sie unter:

Please use the following link for more information on all talks planned in this semester:

https://www.inf.uni-hamburg.de/home/kolloquium/sose18.html

 

Im Namen des Kolloquium-Komitees

On behalf of the colloquium committee

Stephanie Schulte Hemming

Universität Hamburg

 

 

ABSTRACT:

Convolutional highways are based on multiple stacked convolutional layers for feature preprocessing. Like many other convolutional networks convolutional highways are parameterized by numerous hyperparameters that have to be tuned carefully. We introduce an evolutionary algorithm (EA) for optimization of the structure and tuning of hyperparameters of convolutional highways and demonstrate the potential of this optimization setting on the well-known MNIST data set. The EA employs Rechenberg's mutation rate control and a niching mechanism to overcome local optima. An experimental study shows that the EA is capable of evolving convolutional highway networks from scratch with only few evaluations but achieving competitive accuracy.

 

BIO:

Oliver Kramer's research interests are evolutionary optimization and deep learning with applications to real-world domains. He received a PhD from the University of Paderborn, Germany, in 2008. After postdoc stays at the TU Dortmund (Germany), Stanford and Berkeley (USA), he became professor for Computational Intelligence at the Department of Computing Science at the University of Oldenburg in 2011. He is the author of six books on evolutionary computation and machine learning.

  

CONTACT:

Prof. Dr. Mathias Fischer