Colloquium: Evolution of Convolutional Highway Networks

2 July 2018

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.

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.

Monday, 02 July 2018, 17:15, Konrad-Zuse Hörsaal, B-201, Informatics Campus, Stellingen

Speaker: Prof. Dr. Oliver Kramer, Universität Oldenburg.