Transfer Learning Knowledge - Four Application Scenarios of Transfer Learning

Migration learning is to train under a known working condition, and then the trained method (model) can solve the problem under unknown working conditions,
so it includes four aspects

1. Different working conditions: when the working conditions such as constant speed, load and temperature are different, domain shift may occur. For example, a good model trained at a rotational speed of 1730rpm can also satisfy good performance when tested at a rotational speed of 1797rpm

2. Different fault types: The number of fault types in the source domain and the target domain will be unequal. According to the fault types in the source domain and target domain, it can be divided into closed set migration, partial set migration, open set migration and general set migration.

3. Different fault locations: the same equipment, the sensors are installed in different locations

4. Different machines and equipment: the experimental bench is migrated to the real equipment