Transfer Learning
Definition
Transfer learning uses previously learned knowledge for a related new task.
The use of representations or parameters from a source task to improve a related target task.
Simply explained
An image model has already learned many shapes. Therefore, it does not need to learn from scratch to recognize certain defects.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A quality inspector adapts a pretrained image model with a few of their own defect images.
Limits and pitfalls
- Negative transfer, licensing issues, and domain shift can negate benefits; baselines remain necessary.
Origin & history
Transfer learning was researched in the 1990s and became widely practical with pretrained deep networks.
Use in mid-sized companies
- Possible example: A quality inspector adapts a pretrained image model with a few of their own defect images.
Frequently asked questions
What does transfer learning mean?
The use of representations or parameters from a source task to improve a related target task. What should one pay attention to? Negative transfer, licensing issues, and domain shift can negate benefits; baselines remain necessary.
Sources
- How transferable are features in deep neural networks? · Yosinski et al., How transferable are features in deep neural networks?
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: