Self-organizing maps
Hi-index | 0.00 |
This paper presents a method to categorize and supervise time delays by means of Kohonen maps. In detail it will be shown that the evaluation of Weibull and Erlang distributions of the delay data of a company is an essential preprocessing step which enables a Kohonen map to re-evaluate how many classes of delay are coded in the data sets. Further on it can be shown that the knowledge of this topological mapping of time delay classes makes it possible to compute the severity of each delay class by using appropriate weighting factors, such that difficult problems can be remedied first and will not delay the course of the system any further.