In the 1960's the spread of soft-computing techniques gave a great boost to the optimal control of large-complexity systems. The first herald of these methods were the fuzzy logic and set theory, the description of which Zadeh published in 1965 in his well-known book. Thereafter began a rapid improvement of methodology, which still is in progress nowadays. The other two main aspects, the neural networks and evolutionary algorithms, were born after that, which are supposed to mimic the behavior of living organisms and biological processes, in order to be able to control or resolve complex technical processes more easily, quickly and sometimes precisely than before.
In my previous projects dealt with soft-computing techniques and started to implement a primitive soft-computing-based medical decision support system.
During this thesis work I was to create an evolutionary algorithm optimized fuzzy rule-based medical decision support system. In this work a program has been developed, which several evolutionary techniques can be applied for. The developed medical decision support software is capable to diagnose diabetes insipidus with chemical urine analysis.