Transfer learning across domains is primarily used in the areas of artificial intelligence and digital transformation. The term describes a method in which knowledge that has been learnt by artificial intelligence in a specific area (domain) is successfully applied to another area.
Imagine that a computer program has learnt to recognise animals from many photos. Now it needs to identify plants. During transfer learning via domains, the program uses the knowledge it has already learned about shapes and colours from the animal domain to identify the plants more quickly and with fewer new examples.
For companies, this means that instead of starting from scratch for every new problem, existing AI knowledge can be used to speed up solutions and save costs. In practice, this can be used in the area of Industry 4.0, for example, when a machine is first trained for the quality control of car parts and then switches to checking other products without much effort.
Transfer learning across domains ensures that artificial intelligence can be used more flexibly, efficiently and economically.
Deutsch 
English (UK)