Nonlinear mapping of interval vectors by neural networks
K. Kwon, Hisao Ishibuchi, Hideaki Tanaka · 2005
Three approaches are proposed to the learning of neural networks that realize nonlinear mappings of interval vectors. In the proposed approaches, training data for the learning of neural networks are the pairs of interval input vectors and interval target vectors. The first approach is a direct application of the standard backpropagation algorithm with a pre-processor of the training data. The second approach is an extension of the backpropagation algorithm to the case of interval input-output data. The last approach is an extension of the second approach to neural networks with interval weights. These approaches are compared with one another by computer simulations.