New backpropagation algorithm using quadratic potential functions, and an experiment on isolated word recognition

This paper presents a new algorithm to train multilayered perceptrons, u sing q uadr atical potential functions. This new algorithm is compared in an isolated word recognition task, with the back propagation algorithm that uses linear combinations of the inputs. Some pattern recognition techniques are also used to reduce the dimensinality of the input pattern in order to reduce the computational burden of the training and the recognition. The algorithm that uses quadratic potential functions yields better results in the recognition task.