Neural Networks Approach to the Random Walk Dilemma of Financial Time Series
Applied Intelligence
A New Intelligent System Methodology for Time Series Forecasting with Artificial Neural Networks
Neural Processing Letters
Hi-index | 0.00 |
In this work we present the dilation-erosion-linear perceptron (DELP) for financial prediction. It is composed of morphological operators under context of lattice theory and a linear operator. A gradient-based method is presented to design the proposed DELP (learning process). Also, it is included an automatic phase fix procedure to adjust time phase distortions observed in financial phenomena. Furthermore, an experimental analysis is conducted with the proposed model using the Bovespa Index, where five well-known performance metrics and an evaluation function are used to assess the prediction performance.