Loss Function
Definition
A loss function measures how much a model deviates from its training goal.
A function that expresses the deviation between model behavior and training goal as a value to be minimized.
Simply explained
It gives the model an error number. During learning, the model tries to make this number smaller.
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 failure model weights rare, costly false alarms via a professionally justified cost function.
Limits and pitfalls
- Low training loss does not guarantee generalization or business impact; scaling and regularization influence optimization.
Origin & history
Loss functions originate from statistics and optimization and form the basis of many learning methods.
Use in mid-sized companies
- Possible application: A failure model weights rare, costly false alarms via a professionally justified cost function.
Frequently asked questions
What does loss function mean?
A function that expresses the deviation between model behavior and training target as a value to be minimized. What should be considered? Low training loss does not guarantee generalization or business impact; scaling and regularization influence optimization.
Sources
- Deep Learning, Numerical Computation · Goodfellow et al., Deep Learning, Numerical Computation
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: