SqrtLink<T>
Models & Types · AiDotNet.LinkFunctions
Square root link function: g(μ) = √μ.
Implements: ILinkFunction<T>
For Beginners
Use this when:
- You have count data but want to moderate extreme predictions
- The log link produces predictions that are too extreme
The square root link is gentler than the log link:
- log(100) = 4.6, so exp(5) = 148
- sqrt(100) = 10, so 10² = 100
This means changes in the linear predictor have a more moderate effect on predictions.
How It Works
The square root link is often used with Poisson count data as an alternative to the log link. It provides variance stabilization.
Constructors
| Constructor | Summary |
|---|---|
SqrtLink() |
Properties
| Property | Summary |
|---|---|
bool IsCanonical { get; } |
|
string Name { get; } |
Methods
| Method | Summary |
|---|---|
T InverseLink(T eta) |
|
T InverseLinkDerivative(T eta) |
|
T Link(T mu) |
|
T LinkDerivative(T mu) |
|
T Variance(T mu) |