This paper shows how to use the Kohonen algorithm to represent multidimensional data, by exploiting the self-organizing property. It is possible to get such maps as well for quantitative variables as for qualitative ones, or for a mixing of both. The contents of the paper come from various works by SAMOS-MATISSE members, in particular by E. de Bodt, B. Girard, P. Letrémy, S. Ibbou, P. Rousset. ...

Acoustic emission-based techniques are promising for nondestructive inspection of mechanical systems. For reliable automatic fault monitoring, it is important to identify the transient crack-related signals in the presence of strong timevarying noise and other interference. In this paper we propose the application of the Kohonen network for this purpose. The principal components of the short-ti...

A new family of self-organizing maps, the Winner-Relaxing Kohonen Algorithm, is introduced as a generalization of a variant given by Kohonen in 1991. The magnification behaviour is calculated analytically. For the original variant a magnification exponent of 4/7 is derived; the generalized version allows to steer the magnification in the wide range from exponent 1/2 to 1 in the one-dimensional ...

We present a Self-Organizing Kohonen Neural Network for quantizing colour graphics images. The network is compared with existing algorithmic methods for colour quantization. It is shown experimentally that, by adjusting a quality factor, the network can produce images of much greater quality with longer running times, or slightly better quality with shorter running times than the existing metho...

In a recent paper, T. Kohonen and P. Somervuo have shown that self-organizing maps (SOMs) are not restricted to numerical data. They can also be defined for symbol strings, provided that one defines an average function for strings and that the adaptation process is performed off-line (batch). In this paper, we present two different methods for computing averages of strings, as well as an on-lin...

A system based on a neural network framework is considered. We used two neural networks, an Elman network [1][2] and a Kohonen (concurrent) network [3], for a categorization task. The input of the system are objects derived from three general prototypes: circle, square, polygon. We varied the size and orientation of the objects in a continuous way. The system is trained using a new algorithm, b...

Investment decision-making is modeled by means of a Kohonen neural net, where neurons represent firms. This is done in order to model investments in novel fields of economic activity, that according to this model are carried out when firms recognize the emergence of a new technological pattern. Combination of the equations of Kohonen model neuron with macroeconomic relationships yields disaggre...

Vector Quantization can be considered as a data compression technique. In the last few years, vector quantization has been increasingly applied to reduce problem complexity like pattern recognition. In speech recognition, discrete systems are developed to build up real-time systems. This paper presents original results by comparing the KMeans and the Kohonen approaches on the same recognition p...

This paper explores the potential for utilising specialised hardware for the implementation of the Kohonen model and its derivative models. The adaptation periods of these algorithms is potentially protracted and computationally expensive; especially for the Adaptive Kohonen model in real-world, on-line applications. This paper analyses these models and highlights inherent parallelism. This ana...

Action selection is a central issue in the design of behavior-based control architectures for autonomous mobile robots. This paper presents an action selection framework based on an assemblage of self-organizing neural networks called Cooperative Extended Kohonen Maps. This framework encapsulates two features that significantly enhance a robot’s action selection capability: self-organization in...