STATIC AND DYNAMIC PREPROCESSING METHODS IN NEURAL NETWORKS

Preprocessing is recognized as an important tool in modeling, particul arly when the data or underlying physical process involves complex non linear dynamical interactions. This paper will give a review of prepro cessing methods used in linear and nonlinear models. The problem of st atic preprocessing will be considered first, where no dependence on ti me between the input vectors is assumed. Then, dynamic preprocessing m ethods which involve the modification of time-dependent input values b efore they are used in the linear or nonlinear models will be consider ed. Furthermore, the problem of an insufficient number of input vector s is considered. It is shown that one way in which this problem can be overcome is by expanding the weight vector in terms of the available input vectors. Finally, a new problem which involves both cases of: (1 ) transformation of input vectors; and (2) insufficient number of inpu t vectors is considered. It is shown how a combination of the techniqu es used to solve the individual problems can be combined to solve this composite problem. Some open issues in this type of preprocessing met hods are discussed.