A study and development of Bayesian exemplar based models

People think and reason about situations using past experience. Often, pastexperience consists of stereotypes and exemplars that depict common situations.In order for a computer to reason in a similar way, the identification andrepresentation of exemplars is required. This research aims to investigate anddevelop a new model, called FReBE, that uses Family Resemblance andBayesian networks for developing an Exemplar based model.In the thesis, the broad area of research is stated and the areas ofbackground work to be studied are identified and reviewed. An exemplar basedmodel based on family resemblance, clustering and Bayesian networks isdeveloped and implemented. An empirical evaluation of the model is carried outusing fifteen well-known benchmark datasets, such as Breast Cancer, Monks,and Heart Disease. The data sets selected include those with discrete attributes,numerical attributes and even those with missing values. The thesis includes acritical comparison with related systems such as PROTOS, PEBM andAUTOCLASS.The results show that: (a) the model is able to identify exemplars that leadto a classification accuracy comparable to published results with other methodson most of the chosen datasets; (b) most of the datasets can be represented wellby a relatively small number of exemplars which are identified by FReBE; (c)FReBE shows that family resemblance can be used as a principle for findinggood exemplars.

Level dc:type.qualificationlevel

Doctoral (Level 8)

Year dc:date.issued

2008

Author dc:creator

• Wu, J

Language dc:language

en

Identifier

oai:salford-repository.worktribe.com:1336924

OAI identifier oai:identifier

oai:salford-repository.worktribe.com:1336924