Aplicación del modelo neurodifuso ANFIS vs redes neuronales, al problema predictivo de caudales medios mensuales del río Bogotá en Villapinzón

Abstract

This paper shows the results in the exploration of the benefits of the implementation of neuro-fuzzy ANFIS model and neural networks for the prediction of monthly mean flows in the basin of Bogota River in Villapinzón. The ANFIS model is developed, implemented and the performance of six models is evaluated bychanging entries number, number and type of fuzzy sets (membership functions), which are the basic parameters of the ANFIS model. The results are compared with those obtained with multilayer perceptron neural networks.

Keywords

• ANFIS
• Artificial intelligence
• Neural networks.