Class VariablePointOp_Rprop
Inherited Members
Namespace: Microsoft.ML.Probabilistic.Factors
Assembly: Microsoft.ML.Probabilistic.dll
Syntax
[FactorMethod(typeof(Clone), "VariablePoint<>", new Type[]{}, Default = true)]
[Buffers(new string[]{"buffer"})]
[Quality(QualityBand.Preview)]
public class VariablePointOp_Rprop : VariablePointOpBaseMethods
Buffer(Gaussian, Gaussian, Gaussian, RpropBufferData)
Declaration
[SkipIfAllUniform]
public static RpropBufferData Buffer(Gaussian use, Gaussian def, Gaussian to_marginal, RpropBufferData buffer)Parameters
| Type | Name | Description |
|---|---|---|
| Gaussian | use | |
| Gaussian | def | |
| Gaussian | to_marginal | |
| RpropBufferData | buffer |
Returns
| Type | Description |
|---|---|
| RpropBufferData |
BufferInit()
Declaration
public static RpropBufferData BufferInit()Returns
| Type | Description |
|---|---|
| RpropBufferData |
MarginalAverageConditional(Gaussian, Gaussian, RpropBufferData, Gaussian)
EP message to marginal.
Declaration
public static Gaussian MarginalAverageConditional(Gaussian use, Gaussian def, RpropBufferData buffer, Gaussian result)Parameters
| Type | Name | Description |
|---|---|---|
| Gaussian | use | Incoming message from |
| Gaussian | def | Incoming message from |
| RpropBufferData | buffer | Buffer |
| Gaussian | result | Modified to contain the outgoing message. |
Returns
| Type | Description |
|---|---|
| Gaussian |
|
Remarks
The outgoing message is a distribution matching the moments of marginal as the random arguments are varied. The formula is proj[p(marginal) sum_(use,def) p(use,def) factor(use,def,marginal)]/p(marginal).