Machine learning: What is a model?
A model in the sense of machine learning describes a process that calculates a new insight from input values in a specific format (texts, numbers, vectors). The difference to the “algorithm” is that the model depends on several internal parameters, which must first be optimally “adjusted” with a training data set. In addition to the parameters that are fine-tuned by training, there are also configuration parameters that are not automatically adjusted by training, but which still ensure an adjustment of the model and thus its performance. These parameters are referred to as hyper-parameters. And the search for optimal such parameters is called hyper-parameter tuning.
