Real-Time Forecasting by Bio-Inspired Models

摘要

In recent years, bio-inspired methods for problem solving, such as Artificial Neural Networks (ANNs) or Genetic and Evolutionary Algorithms (GEAs), have gained an increasing acceptance as alternative approaches for forecasting, due to advantages such as nonlinear learning and adaptive search. The present work reports the use of these techniques for Real-Time Forecasting (RTF), where there is a need for an autonomous system capable of fast replies. Comparisons among bio-inspired and conventional approaches (e.g., Exponential Smoothing), revealed better forecasting performances for the evolutionary and connectionist models.