Implementation of connectionists systems for trade centers energy consumption forecasting

The problem of energy consumption forecasting in trade centers in Holland and Belgium is considered in this paper. The forecast frameworks based on connectionists systems is suggested here. Paper presents the general description of the framework and the list of specific features connected with observed domain. There are several connectionists systems are implemented to solve forecast problem. They are based on average model, linear regression, neural networks. Models were tested with different set of input variables. In conclusion results and summary were presented.