RSOM Algorithm for Radar Target Recogniton
Lefeng Zhang, Huapeng Yu · 2006
A number of general neural networks have several drawbacks such as how to decide their structures and scales, how to design their self-learning procedure and how to cope with a bulk of computation for the case of large data set classification and complex patterns recognition. In order to solve these problems, this paper proposes the RSOM tree classifier based on discrimination criterion approach. The RSOM tree classifier is composed of topology-preserved sub-SOM nets, and its scales is determined by discrimination criterion. The main advantage of this new neural network is that it adjusts structure and scale automatically with the large training data set, therefore, it maps the training data set very well. This makes it achieve high right classification rate in radar ship target recognition. The experiments in the end are very good proofs for this new network