Knowledge transfer between models (Glossary)

The term knowledge transfer between models comes from the fields of artificial intelligence, big data, smart data and automation. It is about how an AI model that has already been trained can pass on its knowledge to another model so that the latter learns faster and more efficiently.

Imagine that an AI model has already learnt to recognise cats in images very well. Instead of training a second model from scratch, you use the experience of the first model. This saves time, energy and a lot of computing power. The process works in a similar way to humans: If you know how to ride a bicycle, it is easier to learn how to ride a motorbike because many of the principles are similar.

In companies, knowledge transfer between models enables new tasks to be solved more quickly. For example, a model that understands speech can help to train another system to recognise speech in different dialects.

Thanks to knowledge transfer between models, projects in these categories can be organised more efficiently because existing knowledge is used sensibly instead of starting from scratch.