|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Object blog.AbstractCondProbDistrib blog.distrib.LinearGaussian
public class LinearGaussian
Multivariate Gaussian distribution whose mean is an affine function of some vector-valued parents. Specifically, the distribution is Normal(mu + W v, Sigma). Here v is a column vector formed by stacking all the parent vectors on top of each other; we'll call its dimension c. Then mu is a 1xd mean vector, W is a dxc regression matrix, and Sigma is a dxd covariance matrix. Note that the covariance of this distribution does not depend on the parents.
In an input file, this CPD takes three parameters: mu, W, Sigma.
Constructor Summary | |
---|---|
LinearGaussian(java.util.List params)
Takes a parameter list whose first element is a 1xd mean matrix, second element is a dxc regression matrix (where c is the sum of the dimensions of the parent vectors), and third element is a dxd covariance matrix. |
Method Summary | |
---|---|
double |
getProb(java.util.List args,
java.lang.Object value)
For a discrete distribution, returns the conditional probability of childValue given the argument values args . |
java.lang.Object |
sampleVal(java.util.List args,
Type childType)
Samples a value according to this CPD given the args
. |
Methods inherited from class blog.AbstractCondProbDistrib |
---|
getLogProb, toString |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait |
Constructor Detail |
---|
public LinearGaussian(java.util.List params)
Method Detail |
---|
public double getProb(java.util.List args, java.lang.Object value)
CondProbDistrib
childValue
given the argument values args
.
For a continuous distribution, returns the conditional probability
density at childValue
.
public java.lang.Object sampleVal(java.util.List args, Type childType)
CondProbDistrib
args
. The object returned should be of the specified type.
|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |