PBNT.Inference.EnumerationEngine Class Reference

Inheritance diagram for PBNT.Inference.EnumerationEngine:

Collaboration diagram for PBNT.Inference.EnumerationEngine:

Public Member Functions
def	marginal (self, nodes)
Public Member Functions inherited from PBNT.Inference.InferenceEngine
def	__init__ (self, bnet)
def	marginal (self)

Private Member Functions
def	__enumerate_all (self, node, value)
def	__initialize (self, nonEvidence)
def	__next_state (self, nonEvidence)
def	__probability (self, state)

Additional Inherited Members
Public Attributes inherited from PBNT.Inference.InferenceEngine
	bnet
	evidence

Detailed Description

Enumeration Engine uses an unoptimized fully enumerate brute force method to compute the marginal of a query.  It also uses the standard constructor, init_evidence, and change_evidence methods.  In this engine, we use a hack.  We have to check and see if the variable is unobserved.  If it is not, then we know that the probability of that value is automatically 1.  We use this hack, because in order to do it properly, a table of likelihoods that incorporates the evidence would have to be constructed, this is very costly.

Member Function Documentation

def PBNT.Inference.EnumerationEngine.__enumerate_all (   self,   node,   value  )

private

We are going to iterate through all values of all non-evidence nodes. For each state of the evidence we sum the probability of that state by the probabilities of all other states.

---

The documentation for this class was generated from the following file:

• PBNT/Inference.py