One of the most common tools used by designers of automated recognition systems to obtain better results is to utilize data normalization. There are many types of data normalization. Ideally a system designer wants the same range of values for each input feature in order to minimize bias within the neural network for one feature over another. Data normalization can also speed up training time by starting the training process for each feature within the same scale. Data normalization is especially useful for modeling applications where the inputs are generally on widely different scales. The authors present some of the more common data normalization techniques in this chapter, beginning with statistical or Z-score normalization.
Online access to SPIE eBooks is limited to subscribing institutions.