Package com.bayesserver.analysis

Class ValueOfInformation

java.lang.Object

com.bayesserver.analysis.ValueOfInformation

public final class ValueOfInformation
extends Object

Contains methods to determine what new evidence is most likely to reduce the uncertainty of a variable.

Value of information is useful when you have a particular variable of interest (the hypothesis variable) and you want to know what additional evidence is most likely to reduce the uncertainty of that variable.

For example in a medical troubleshooter, you may wish to know which questions to ask or which medical checks to carry out first, in order to determine whether the patient has a particular disease or not.

Method Summary

Modifier and Type	Method	Description
static ValueOfInformationOutput	calculate(VariableContext hypothesis, List<VariableContext> testVariables, Evidence evidence, InferenceFactory factory, ValueOfInformationOptions options)	Calculates value of information, which can be used to determine which variables are most likely to reduce the uncertainty of a particular variable.
static ValueOfInformationOutput	calculate(Variable hypothesis, List<Variable> testVariables, Evidence evidence, InferenceFactory factory, ValueOfInformationOptions options)	Calculates value of information, which can be used to determine which variables are most likely to reduce the uncertainty of a particular variable.

Methods inherited from class java.lang.Object

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

Method Detail

calculate

public static ValueOfInformationOutput calculate(Variable hypothesis,
                                                 List<Variable> testVariables,
                                                 Evidence evidence,
                                                 InferenceFactory factory,
                                                 ValueOfInformationOptions options)
                                          throws InconsistentEvidenceException

Calculates value of information, which can be used to determine which variables are most likely to reduce the uncertainty of a particular variable. See ValueOfInformation for more information.

Parameters:

hypothesis - The variable whose uncertaintly you are trying to reduce. E.g. for fault diagnosis / troubleshooting.

testVariables - The variables you wish to consider in order to reduce the uncertainty of the hypothesis variable.

evidence - Any evidence, which can be null.

factory - Determines the type of inference engine to use during the calculations.

options - Options affecting how value of information is calculated.

Returns:

Information about how the [testVariables] influence the [hypothesis] variable.

Throws:

InconsistentEvidenceException - Raised when the hypothesis distribution cannot be calculated given the evidence.

calculate

public static ValueOfInformationOutput calculate(VariableContext hypothesis,
                                                 List<VariableContext> testVariables,
                                                 Evidence evidence,
                                                 InferenceFactory factory,
                                                 ValueOfInformationOptions options)
                                          throws InconsistentEvidenceException

Calculates value of information, which can be used to determine which variables are most likely to reduce the uncertainty of a particular variable. See ValueOfInformation for more information.

Parameters:

hypothesis - The variable (and time for temporal variables) whose uncertaintly you are trying to reduce. E.g. for fault diagnosis / troubleshooting.

testVariables - The variables (and associated times for temporal variables) you wish to consider in order to reduce the uncertainty of the hypothesis variable.

evidence - Any evidence, which can be null.

factory - Determines the type of inference engine to use during the calculations.

options - Options affecting how value of information is calculated.

Returns:

Information about how the [testVariables] influence the [hypothesis] variable.

Throws:

InconsistentEvidenceException - Raised when the hypothesis distribution cannot be calculated given the evidence.