CS-587

Neural Networks and Learning of Hierarchical Representation

Postgraduate

The scientific activity of the last decade revealed many new directions and highly successful extensions of Neural Networks towards learning data representations for various perception systems. Representations of this kind are composed of many layers of nonlinear calculations (multilayer architectures) and are based on classic artificial neural networks. In recent years it has become evident that learning such multilayered representations can contribute to a significant improvement in perception systems performance. The purpose of this course is to present an introduction to artificial neural networks and in learning hierarchical representations based on those network structures. The course will focus on architectures, methodologies and algorithms, and will also include laboratory exercises.

6

CS-217, CS-119 or Μ-105