Structural learning of neural networks for forecasting stock prices

Abstract

Generally, a neural network spends much computation time and cost in forecasting the value and movement of a stock. The reason is because a neural network requires exponential time in computation according to the number of units in a hidden layer. The objective of the paper is to optimally build a neural network through structurally learning. The results enable us to reduce the computational time and cost as well as to understand the structure more easily. In the paper the method is employed in forecasting the price movement of a stock. The optimization of the network by the structured learning is evaluated based on its real use. © Springer-Verlag Berlin Heidelberg 2006.