The thesis commences with a review of basic Static Bayesian Networks (SBNs) and describes some of the methods of probabilistic inference from SBNs, including Pearl's causal tree methodology and the more general NP-complete clique tree methodology due to Lauritzen and Speigelhalter. The use of SBNs in pattern classification will then be described with special consideration given to the use of elementary forms of SBNs including the Naive Bayes and the Tree Augmented Naїve Bayes (TAN) classifiers. The possibility of applying boosting and other ensemble methods to these elementary forms in order to obtain superior performance will then be explored. The discussion will conclude by investigating the suitability of correlation measures in computing the TAN as an initial step in adapting it to solve the forecasting problem. Attention will then be turned to the use of Bayesian Networks (BN) in forecasting. Numerous techniques have been developed to create accurate forecasting models and the BN approach, with a time element incorporated, is among the more successful of these. The time element will be introduced by means of a series of SBNs, acting as time-slices to create a Dynamic Bayesian Network (DBN), to solve the forecasting problem. The suitably modified TAN combined with the Pearl causal tree, now called the TAN-Pearl Network (TPN), will form the basis of the DBN. The objective of the forecasting problem will be to minimize overall error, computed from the differences between the computed beliefs at each time-slice, and those of the actual events. The final major investigation concerns the application of boosting to regression. It draws upon the parallel between time-slices in a DBN and instances in regression analysis and it will be shown that the accuracy of the time-slices, and therefore that of the DBN as a whole, can be improved through boosting.