Machine Learning

There are three methods of machine learning: Supervised machine learning is based on the concept that a machine learns from data and transfers this knowledge to other data and situations after the learning phase has ended. Algorithms build a statistical model based on training data during machine learning. This means that the model does not simply remember the data with which it was trained, but it recognizes patterns and regularities in the training data. Thus, the model can possibly learn something new from data or apply learned (training data) to other data (current data), e.g., make predictions. Before a model is considered “trained” (trained model) and applied to other data, the patterns and regularities are evaluated with the help of test data. If the quality of the model (model quality) is considered sufficient, then the trained model is applied. The results of the trained model should also be continuously checked to recognize when, for example, old patterns no longer apply meaningfully to the new data and thus a new training of the model (re-training) should be initiated.