A Coyote Optimization Algorithm-trained Dendritic Neuron Model for Prediction Tasks

Time series prediction is always a hot topic. Many artificial neural network-based models are continuously proposed to improve the predicting performance. However, in practical situations, the large scale of training data and the complex structure of network often lead to slow training speed and unsatisfactory result. Thus, it is reasonable to use a suitable algorithm to train a well-structured network for such an issue. In this paper, we use the coyote optimization algorithm to train a dendritic neuron and propose the COADNM to tackle time series prediction problems. Three datasets are used to verify the performance of the proposed model, meanwhile, four other neural network-based models are employed to compare with COADNM. The experiment result shows that our model outperforms all its competitors via different evaluation criteria.