Inference-Based Bayesian Network For Threatened Tree Species

Vol. 6 No. 3

Year : 2011

Issue : Feb-Apr

Title  : Inference-Based Bayesian Network For Threatened Tree Species 

Author Name  : Behailu Getachew 

Synopsis  : 

In  the  recent  years,  Bayesian  Net  (BN)  is  an  increasingly  popular  formalism  for reasoning and  decision-making  in  problems  that   involve  uncertainty  and probabilistic reasoning. Bayesian Net gives the new advancements and innovations in Threatened  Tree Species   and   also   fulfilling   the   barriers   between   Knowledge Representation   and Artificial   intelligence.  The  species   degradation   calls   for  uncertainty  management  technique  that  aims  at  developing  knowledge  base  and  evaluating  the  status  of  the  species.  BN is a machine learning tool to provide an opportunity for predicting the species uncertainties in an Artificial intelligence system. 

This paper reports a comparison of BN-based and subjective (elicit-based) prediction modeling and hence, relevant inference examples have been taken from the generated conditional  probability  table  (CPT)  so  as  to  define  decision  making  of  the  problem domain. The results demonstrate the performance prediction and inference BN model for   biological   variables   or   knowledge   base   via   Three-Phase-Dependency Analysis algorithm.