com.numericalmethod.suanshu.stats.regression.linear.glm.quasi

Class GeneralizedLinearModelQuasiFamily

java.lang.Object

com.numericalmethod.suanshu.stats.regression.linear.glm.quasi.GeneralizedLinearModelQuasiFamily

All Implemented Interfaces:

java.io.Serializable

public class GeneralizedLinearModelQuasiFamily
extends java.lang.Object
implements java.io.Serializable

GLM for the quasi-families.

In order to construct a likelihood function it is usually necessary to posit a probabilistic mechanism specifying, for a range of parameter values, the probabilities of all relevant samples that might possibly have been observed. Such a specification implies the knowledge of the mechanism by which the data were generated or substantial experience of similar data from previous experiments. Often, this knowledge is not available.

We may, however, be able to specify the range of possible response values and past experience with similar data is usually sufficient to specify, in a qualitative fashion, a few additional characteristic features of the data. From these characteristics, we may construct a quasi-likelihood function.

Note that AIC is not computed for the quasi-GLM because there is no 'real' likelihood function.

See Also:

"P. J. MacCullagh and J. A. Nelder, "An algorithm for fitting generalized linear models," in Generalized Linear Models, 2nd ed. Chapter 9.", Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
Beta	beta the GLM coefficients β^ statistics
QuasiGlmProblem	problem the quasi- generalized linear regression problem to be solved
Residuals	residuals the residual analysis of this quasi GLM regression

Constructor Summary

Constructors

Constructor and Description
GeneralizedLinearModelQuasiFamily(QuasiGlmProblem problem) Construct a GeneralizedLinearModelQuasiFamily instance.

Method Summary

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

problem

public final QuasiGlmProblem problem

the quasi- generalized linear regression problem to be solved

beta

public final Beta beta

the GLM coefficients β^ statistics

residuals

public final Residuals residuals

the residual analysis of this quasi GLM regression

Constructor Detail

GeneralizedLinearModelQuasiFamily

public GeneralizedLinearModelQuasiFamily(QuasiGlmProblem problem)

Construct a GeneralizedLinearModelQuasiFamily instance.

Parameters:

problem - the quasi generalized linear regression problem to be solved