I have generated a multiple-regression equation/model from the data.

Parameter A in the data varies from $0.1$ – $10$.

Parameter B in the data varies from $100$ – $1800$.

Parameter C in the data varies from $0.1$ – $0.5$.

The model is,

$y = c + \log A + \frac{1000}{B} + \log C$

Is this regression equation valid outside the range of the above parameters?

**Is there any method to find out in what range of A, B and C the above regression model is applicable/valid?**

## Best Answer

Mathematically, the equation would be

validwhen $A>0, C>0$ and $B\neq 0$, you can substitute and find some $y$ for any given $A,B,C$ satisfying these conditions.Estimating outside the ranges is extrapolating and since you do not have data to validate those ranges, it'd be nearly impossible to know if your model does a good job there; except the domain know-how you can apply on the found target values.