Machine learning describes computer models, which can learn pattern-recognition. It uses large amounts of data, which are differentiated according to the desired patterns. The model recognizes the connection of the data and is able to give an estimate. After the training, the precision of the model is tested and afterwards adjusted through new data until the model achieves desired results. This process can be used repeatedly in succession to enable more and more accurate pattern recognition. The training process in turn is done by trained models, thus becoming more and more efficient and accurate. Without machine learning it would be logistically impossible to handle the amount of data (big data) being produced. These models are already more accurate and far more efficient in recognizing patterns and diagnostics than humans. Once the model is created, it can be continuously improved.