Applications of recent learning and soft computing techniques to humanistic systems

Although modern computer is the most revolutionary and most powerful tool developed in the twenty's century, it is almost useless for the application of this tool to the not well defined humanistic systems such as politics, law, or even the many hour-to-hour small decisions we all make routinely and daily. This is in spite of the fact that the human action of the cognitive band, which is of the order of seconds, is much slower than the speed of the modern computer. In this talk, we shall first examine the basic differences between the scientific systems and the humanistic systems. Then, based on these resulting differences, we shall propose a fuzzy-neural combined approach, namely, the fuzzy adaptive network (FAN), to model and to investigate the humanistic systems on modern computer. We shall show that this FAN network; because of its linguistic representation and learning abilities, possesses the required ingredients to be an effective modeling and investigation approach for the humanistic systems. Finally, the fuzzy support vector machines (FSVM) approach is introduced and the advantages of this second generation learning approach are compared with the fuzzy neural network. To illustrate the effectiveness of the proposed approaches, several actual humanistic systems are modeled and investigated.