Surrogate Model
Definition
A surrogate model approximates the behavior of a complex or hard-to-access system with less effort.
A surrogate model approximates an expensive, slow, or not directly accessible system to perform predictions, optimization, or sensitivity analyses more cost-effectively.
Simply explained
Like a simplified stand-in: Instead of querying the expensive original every time, you use an approximation for calculations and comparisons.
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 manufacturer trains a Gaussian process on simulation results to optimize parameters faster.
Limits and pitfalls
- Outside the training domain, the approximation can fail; uncertainty and validity range must be documented.
Origin & history
Response surface methods and Kriging have been established in statistics and engineering; machine learning expanded the model classes.
Use in mid-sized companies
- Possible example: A manufacturer trains a Gaussian process on simulation results to optimize parameters faster.
Frequently asked questions
What does surrogate model mean?
A surrogate model approximates an expensive, slow, or not directly accessible system to perform predictions, optimization, or sensitivity analyses more cost-effectively. What should you watch out for? Outside the training domain, the approximation can fail; uncertainty and validity range must be documented.
Sources
- Design and Analysis of Computer Experiments · Sacks et al., Design and Analysis of Computer Experiments
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: