Application of rough set theory for determining the significant inputs of an ANN [power trading calculations]

In this paper, the authors show a real life application of Pawlak's rough set theory and neural networks in the area of power trading. European power systems are closely interconnected with each other, resulting in unexpected loop flows. This phenomena and the lack of commercially viable information make it very difficult for power traders to trade successfully with power. Rough set theory was used to determine the significant inputs of a neural network that could be applied in trading activity. In this paper, the authors also present a general optimization technique and a real application of the neural nets in power systems.