object oriented modeling, numerical clustering, BETA

This paper deals with object oriented modeling in cluster analysis based on the Mjolner BETA System and the BETA object oriented language. The model is designed in layers, which enables more flexibility and clear understanding. Object oriented modeling enables to create a model that is more natural and is easy to extend or change its functionality. These features predetermine such a model for making experiments. Cluster analysis has many methods and techniques for helping to solve classification problems. Object oriented model enables to exploit them in a more complex way and thus receive interesting results. We use this model on medical data for making a diagnosis. With the benefit we used rich means for composition, virtual mechanism and persistence of the Mjolner BETA System. Persistence allows to split the whole application into rather simple parts and to store the results for the next usage.

This paper deals with object oriented modeling in cluster analysis based on the Mjolner BETA System and the BETA object oriented language. The model is designed in layers, which enables more flexibility and clear understanding. Object oriented modeling enables to create a model that is more natural and is easy to extend or change its functionality. These features predetermine such a model for making experiments. Cluster analysis has many methods and techniques for helping to solve classification problems. Object oriented model enables to exploit them in a more complex way and thus receive interesting results. We use this model on medical data for making a diagnosis. With the benefit we used rich means for composition, virtual mechanism and persistence of the Mjolner BETA System. Persistence allows to split the whole application into rather simple parts and to store the results for the next usage.