Chaotic Time Series Forecasting using Emotional Learning-Based Neural Networks
Amani Fatemeh, Roya Amjadifard · 2019
This paper has investigated a new model based on the brain emotional learning to forecast chaotic series. The proposed model is based on the Moren model which is a model of the emotional part of the brain of mammals. In this paper, we have extended the Moren model based on the actual inputs to the thalamus of mammalian brain, i.e., the number of entries to the thalamus is considered to be more than one. A better performance in chaotic time series forecasting has been achieved in simulations. The results show that in comparison with other methods such as neural network based models and fuzzy neural models, the proposed model has less MSE in both short-term and long-term forecasting for the chaotic time series.