Scaled hyperparameter optimisation is a term from the world of Artificial Intelligence, Big Data and automation. Hyperparameters are settings in an AI model that have a large influence on how well the model performs. Examples of hyperparameters include the learning rate or the size of the data with which an AI is trained.
„Scaled“ hyperparameter optimisation means that not just a few settings are tested, but many different possibilities are tried out automatically and simultaneously – often on multiple computers or with the help of cloud services. This saves a lot of time and can significantly improve the quality of AI models.
Suppose an online shop wants to use Artificial Intelligence to predict which products will sell particularly well. To get the best prediction, the AI system must be optimally configured. With scaled hyperparameter optimisation, hundreds of variations are automatically tested in a short time until the best setting is found – and this without laborious manual work.
This method allows companies to develop powerful and reliable AI models more quickly, thereby making their processes more efficient.
Deutsch 
English (UK)