Microsoft Research, Cambridge
Infer.NET Gaussian Process kernels
Classes
| Class | Description | |
|---|---|---|
| ARD |
Automatic Relevance Determination Kernel
| |
| KernelFactory |
Kernel Factory - singleton class. It maintains a list
of Kernel names and their types
| |
| KernelFunction |
Base class for all kernel functions
| |
| LinearKernel |
Linear kernel
| |
| NNKernel |
Neural Net kernel
| |
| SquaredExponential |
Squared Exponential kernel function: k(x,y) = exp(-0.5*(x-y)^2/exp(2*logLength))
| |
| SummationKernel |
Summation kernel. This provides the management layer for adding
together different kernels
| |
| WhiteNoise |
White noise kernel function. This can be added to other kernels using the
SummationKernel class to provide measurement noise
|
Interfaces
| Interface | Description | |
|---|---|---|
| IKernelFunction |
Interface for all Kernel functions
| |
| IKernelFunctionWithParams |
Interface for Kernel functions with parameters
|