Class RelevanceTreeInference

An exact probabilistic inference algorithm for Bayesian networks and Dynamic Bayesian networks, that can compute multiple distributions more efficiently than the VariableEliminationInference algorithm.

Inheritance

System.Object

RelevanceTreeInference

Implements

IInference

System.IDisposable

Inherited Members

System.Object.Equals(System.Object)

System.Object.Equals(System.Object, System.Object)

System.Object.GetHashCode()

System.Object.GetType()

System.Object.MemberwiseClone()

System.Object.ReferenceEquals(System.Object, System.Object)

System.Object.ToString()

Namespace: BayesServer.Inference.RelevanceTree

Assembly: BayesServer.Inference.dll

Syntax

public sealed class RelevanceTreeInference : IInference, IDisposable

Constructors

RelevanceTreeInference(Network)

Initializes a new instance of the RelevanceTreeInference class, with the target Bayesian network.

Declaration

public RelevanceTreeInference(Network network)

Parameters

Type	Name	Description
Network	network	The target Network.

Exceptions

Type	Condition
System.ArgumentNullException	Raised if network is null.

Properties

BaseEvidence

Optional evidence which can be used to calculate the lift of queries. Defaults to null.

Declaration

public IEvidence BaseEvidence { get; set; }

Property Value

Type	Description
IEvidence

Remarks

When Comparison is true queries are adjusted based on marginals calculated with this evidence or no evidence if this value is null.

See Also

Comparison

Evidence

Represents the evidence, or case data (e.g. row in a database) used in a BayesServer.Inference.IInference.Query.

Declaration

public IEvidence Evidence { get; set; }

Property Value

Type	Description
IEvidence

Remarks

The QueryDistributions are only recalculated by BayesServer.Inference.IInference.Query, not each time evidence is changed.

For more information see IEvidence.

Network

The target Bayesian network.

Declaration

public Network Network { get; }

Property Value

Type	Description
Network

QueryDistributions

Gets or sets the collection of distributions to calculate. See IQueryDistributionCollection.

Declaration

public IQueryDistributionCollection QueryDistributions { get; set; }

Property Value

Type	Description
IQueryDistributionCollection

Remarks

If required, a query distribution collection can be dynamically attached or detached. Also note that individual elements can be enabled or disabled on a per query basis.

QueryFunctions

Gets or sets the collection of functions to evaluate, after QueryDistributions have been calculated.

Declaration

public IQueryFunctionCollection QueryFunctions { get; set; }

Property Value

Type	Description
IQueryFunctionCollection

Remarks

If required, a query function collection can be dynamically attached or detached. Also note that individual elements can be enabled or disabled on a per query basis.

Methods

Dispose()

Stops listening to notifications on the underlying Network.

Declaration

public void Dispose()

Finalize()

Stops listening to notifications on the underlying Network before the RelevanceTreeInference instance is reclaimed by garbage collection.

Declaration

protected void Finalize()

Query(IQueryOptions, IQueryOutput)

Calculates a number of distributions, e.g. P(A) and P(B) given the IEvidence (case data, e.g. row in a database), and if requested the log-likelihood of the data.

Declaration

public void Query(IQueryOptions queryOptions, IQueryOutput queryOutput)

Parameters

Type	Name	Description
IQueryOptions	queryOptions	Options governing which calculations are performed and how.
IQueryOutput	queryOutput	Returns any information, in addition to the IQueryDistributionCollection, that is requested. For example the LogLikelihood.

Remarks

Each time this method is called the IQueryDistributionCollection have their values replaced, acting like buffers.

As well as requesting distributions of the form P(A), P(B) it is also possible to request distributions over a number of variables such as P(A,B).

If CLGaussian distributions are requested that include instantiated discrete variables, an algorithm may legitimately return mean and covariance values equal to NaN for entries corresponding to inconsistent discrete combinations. For example, if a CLGaussian is requested that includes a discrete variable 'Gender' with states 'Male' and 'Female' and evidence is set to 'Female' then any mean and covariance entries that correspond to 'Male' may be set to NaN. NaN values will always be accompanied by a corresponding zero value in the Table.

Exceptions

Type	Condition
InvalidNetworkException	Raised if the network is invalid.
InconsistentEvidenceException	Raised if inconsistent evidence is detected, or underflow/overflow occurs when calculating queries or the conflict measure. Note that log-likelihood calculations do not raise this exception, but instead report -Infinity or +Infinity.

Events

QueryExecuted

The event that is fired when the BayesServer.Inference.IInference.Query method completes.

Declaration

public event EventHandler<QueryExecutedEventArgs> QueryExecuted

Event Type

Type	Description
System.EventHandler<QueryExecutedEventArgs>

QueryExecuting

The event that is fired when the BayesServer.Inference.IInference.Query method begins.

Declaration

public event EventHandler<QueryExecutingEventArgs> QueryExecuting

Event Type

Type	Description
System.EventHandler<QueryExecutingEventArgs>

Implements

IInference

System.IDisposable

See Also

IInference