Power Fault Data Analysis and Visualisation via SOM Neural Networks

J. Martinovič, P. Moravec, J. Dvorský, and V. Snášel (Czech Republic)


WEBSOM, SOM, neural networks, power faults


The data on power faults have been collected in Czech Republic for several years now, but the common frame work which allowed to combine them into one large dataset was completed recently. The next step is to analyze this data and present the underlying knowl edge in such a way, that can be easily understood. In this paper, we will describe the SOM method (and its modification introduced in WEBSOM), based on Ko honen self-organizing neural network, which was al ready successfully used in many areas and is known to capture underlying concepts.

Important Links:

Go Back