Modeling ANNs performance on time series forecasting

Artificial Neural Networks (ANNs) are accurate models that can learn the characteristics of the problem being solved, mainly these are used for classification and regression problems. In this work, the performance of ANNs used in Time Series forecasting is modeled. Modeling the performance of ANNs is specially useful when the algorithm selection problem is being tackled. In order to model different ANNs, only changes in the training parameters are being considered. The influence of the training parameters on the performance is also determined in order to assure that different ANNs are being modeled.