Abstract
Most conventional methods extract knowledge from data in one place. It is, however, often the case where data are generated and stored in two or more places. For extracting fuzzy rules from distributed data, we have proposed a method that transfers only values necessary for the calculation process without collecting all distributed data. But the method assumes that all distributed data have the same attributes. In this paper, we propose a method to extract fuzzy rules from distributed data with different attributes based on the previous method. We illustrate a result for experiments using Iris data by R.A.Fisher.
