Regression
Definition
Regression estimates numerical target values or conditional relationships.
A model class for estimating a numerical target variable or a conditional relationship from explanatory variables.
Simply explained
A model calculates, for example, an estimated energy consumption from temperature and machine runtime.
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 model estimates energy consumption from temperature and machine runtime.
Limits and pitfalls
- Interpreting correlation as causation, not checking assumptions, or allowing data leaks.
Origin & history
Regression was statistically shaped in the 19th century and today includes linear, generalized, and nonlinear variants.
Use in mid-sized companies
- Possible application: A model estimates energy consumption from temperature and machine runtime.
Frequently asked questions
What does regression mean?
A model class for estimating a numerical target variable or a conditional relationship from explanatory variables. What mistakes should be avoided? Interpreting correlation as causation, not checking assumptions, or allowing data leaks.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: