Predictions of the power production for a wind power plant or collection of wind power plants can be improved by using so-called self-adaptive models. These models learn continuously from past experience and are thus able to adjust to changing circumstances, such as seasonal changes, small changes in weather model set-up, or even changes to the numbers or types of wind turbines. This article compares the predictions of Vejr2's neural network model with that of the company's advanced physical model. The neural network model is seen to perform significantly better than the physical model.