Bayesian Networks to Land Planning Knowledge Discovery and Decision Support

This report is for the literature concerning Bayesian Networks, which will be applied to our collaboration project with ACT Planning and Land Authority (ACT-PLA) involving population prediction, suburb development potential, and land supply strategy, etc. Bayesian Network graphically represents a problem domain and is able to integrate multiply data sources as well as conduct bi-directional inferences. So it is widely used in the areas like data mining, knowledge discovery and decision support system. During the literature review, I managed to understand those essential fundamentals of Bayesian Networks including the exact inference algorithms (Variable Elimination, Belief Propagation, and Junction Tree Algorithm) and general steps of constructing Bayesian Networks. Intending to finish the project, more literature review needs to be done next semester. A detailed future research plan is produced at the end of this report. This report is for evaluation of COMP6720, semester 2, 2006 as an equivalent 6-unit course.