
This is an important question for Verbund - Austria's leading electricity company - since it operates 90 hydropower plants with a combined maximum capacity of over 6,000 megawatts and an average annual generation of approx. 22.8 billion kWh.

Hence in its headquarter Verbund runs a prediction system which produces online highly accurate predictions of the available amount of water. An important part of this prediction system relies on neural networks.

The combination of the machine learning framework's high quality neural networks and Mathematica's data import and visualization capabilities have made this technology Verbund's choice to design neural networks for their online prediction system.