Fully Automatic Bayesian Neural Forecaster - NN GC1

This paper combines several techniques to generate a fully data-driven forecasting model. Input selection is performed, without user intervention, by applying chaos theory and Bayesian inference. Afterwards, neural network models are estimated, without cross-validation, relying on data partitioning and Bayesian regularization for complexity control. Automatic data clustering has been used for data partitioning. The proposed forecasting model has been tested with datasets provided by the competition organizers.