Neuroinformatics Group
Universität Bielefeld › Technische Fakultät › NI
Search
Neural Networks
neural network models
Neuroinformatics Group
Universität Bielefeld › Technische Fakultät › NI
Self-organizing maps (SOMs) are a class of unsupervised learning methods that arrange data samples in a low-dimensional, non-linear manifold in a topology-preserving fashion, i.e. neighbored samples will be mapped to neighbored regions in the manifold. Thus they are well-suited for non-linear dimensionality reduction and visualization. 
The Competitive Layer Model (CLM) is a recurrent neural network architecture for dynamic feature binding and relaxation labeling, which was introduced by Helge Ritter in 1990. It can be used for various tasks in perceptual grouping, image segmentation, texture segmentation und classification. The network architecture consists of several topographically ordered layers within which lateral interactions control the binding of features into groups whereas competitve interactions between layers cause the segmentation and labeling of these groups. The stable states of the dynamics can be shown to be minima of a binding energy and implement proper groupings in the sense that each feature is uniquely assigned to a group.
Artificial neural networks try to capture aspects of information processing of biological neural nets in artificial systems. One aim lies in an exploration and testing of hypotheses about the working principles of real neural nets, using simulation models at varying levels of abstraction. A second goal is to exploit attractive properties of neural information processing, such as error tolerance, parallel distributed processing as well as learning ability for technical applications. Besides work in the areas of robotics, computer vision, human machine interfaces and datamining we pursue basic research to topics including stability of recurrent networks, properties of competitive layer networks, neural learning and self-organizing maps.