New Challenges in Neural Computation (NC²)

Workshop of the GI-Fachgruppe Neuronale Netze and the German Neural Networks Society in connection to DAGM 2012, Graz

Workshop program

The duration of the regular talks is limited to 15 min each to allow time for discussions. Please explicitly address open problems/questions/challenges in the presentations!

9:00-10:30	Challenges in Robotics Matthias Rolf, Jochen Steil (Bielefeld University): Explorative learning of right inverse functions: theoretical implications of redundancy Mathias Klingner¹, Sven Hellbach¹, Marika Kästner², Thomas Villmann², Hans-Joachim Böhme¹ (1-University of Applied Sciences Dresden, 2-University of Applied Sciences Mittweida): Modeling Human Movements by Self-Organizing Maps using Adaptive Metrics Jakob Gütschow, Johannes Lohmann, Danil Koryakin, and Martin V. Butz (University of Tübingen): Learning Motor Primitives with Echo State Networks Sven Hellbach, Hans-Joachim Böhme (University of Applied Sciences Dresden): Robotics with the head in the clouds
10:30-11:00	Coffee break
11:00-12:00	Learning non-Euclidean data Emil Eirola¹, Amaury Lendasse¹, Vincent Vandewalle², Christophe Biernacki² (1-Aalto University, 2-University of Lille): Mixture of Gaussians for distance estimation with missing data Thomas Villmann, Marika Kaestner, David Nebel, Martin Riedel (University of Applied Sciences Mittweida): Enhancement Learning in Functional Relevance Learning Vector Quantization Daniela Hofmann, Andrej Gisbrecht, Barbara Hammer (Bielefeld University): Discriminative probabilistic prototype based models in kernel space
12:00-13:00	Keynote talk Prof. Dr. Michel Verleysen (Université catholique de Louvain, Engineering Faculty - Electricity Department): Information theoretic feature selection for high-dimensional data analysis
13:00-14:00	Lunch break
14:00-15:30	Visualization, Interpretation, and Sparsity Marc Strickert (Philipps University Marburg): No Perplexity in Stochastic Neighbor Embedding Alexander Schulz, Andrej Gisbrecht, Kerstin Bunte, and Barbara Hammer (Bielefeld University): How to visualize a classifier? Slobodan Vukanovic, Nicole Carey, Robert Haschke, Helge Ritter (Bielefeld University): Locally Weighted Regression using an Error-based Allocation Strategy Andreas Backhaus, Udo Seiffert (Fraunhofer IFF): Classification in High-dimensional Spectral Data - Precision vs. Interpretability vs. Model Size
15:30-16:00	Coffee break
16:00-17:00	Invariances and Feature Learning Marian Himstedt, Sven Hellbach, Hans-Joachim Böhme (University of Applied Sciences Dresden): Feature extraction from Occupancy Grid Maps using Non-negative Matrix Factorization Markus Lessmann, Rolf Würtz (Ruhr-University Bochum): Learning of Invariant Object Recognition in a Hierarchical Network Jens Hocke, Thomas Martinetz (University of Lübeck): Experience in Training (Deep) Multi-Layer Perceptrons to Classify Digits
17:00-17:20	Discussion: Hot challenges in Neural Computation
17:20-17:30	Nomination of the best presentation award and closing
starting from 17:30	Meeting of the German Neural Network Society and the GI Arbeitskreis Neuronale Netze
19:30-21:30	Welcome reception of DAGM at the old University (venue)

Proceedings

The workshop proceedings will be available online in the series Machine Learning Reports.

Venue

The workshop will take place at the Inffeld Campus of TU-Graz (venue) in room i12, building Inffeldgasse 16b. Take Tram 6 to St.Peter-Schulzentrum, from there it is a 2 min walk.

Registration

Registration to the workshop is available via the DAGM 2012 webpage for a reduced rate or can be done on site. Registration for NC² only is possible; alternatively, the registration for DAGM includes the registration to the workshop.

The workshop is sponsored by the European Neural Networks Society