We investigate new approaches for knowledge discovery from two medical databases. Two different kinds of knowledge, namely rules and causal structures, are learned. Rules capture interesting patterns and regularities in the database. Causal structures represented by Bayesian networks capture the causality relationships among the attributes. We employ advanced evolutionary algorithms for these d...

Bayesian analysis and modeling, in which uncertainties are quantified in terms of probability, offers an alternative approach to understanding in meteorological applications. The underlying principle, practiced in fields like archaeology and geology, is the accumulation of evidence. The approach provides a mathematical rule to update existing beliefs in light of new evidence. It requires the da...

Although Bayesian model averaging is theoretically the optimal method for combining learned models, it has seen very little use in machine learning. In this paper we study its application to combining rule sets, and compare it with bagging and partitioning, two popular but more ad hoc alternatives. Our experiments show that, surprisingly, Bayesian model averaging’s error rates are consistently ...

The Bayesian logic is generally associated to the definition of a prior probabilistic law. Conditional algebra have been investigated by some authors though, but somehow the background framework is still probabilistic and the entire logic is not specified. In this paper, the definition of a Deterministic Bayesian Logic is proposed. This logic is completely independent of any notion of probabili...

In this research, the decision on belief (DOB) approach was employed to analyze and classify the states of uni-variate quality control systems. The concept of DOB and its application in decision making problems were introduced, and then a methodology for modeling a statistical quality control problem by DOB approach was discussed. For this iterative approach, the belief for a system being out-...

In this note we consider the problem of, given a sample, selecting the number of bins in a histogram. A loss function is introduced which reflects the idea that smooth distributions should have fewer bins than rough distributions. A stepwise Bayes rule, based on the Bayesian bootstrap, is found and is shown to be admissible. Some simulation results are presented to show how the rule works in pr...

Abstract I We propose a new framework for surprise-driven learning that can be used for modeling how humans and animals learn in changing environments. It approximates optimal Bayesian learner, but with significantly reduced computational complexity. I This framework consists of two components: (i) a confidence-adjusted surprise measure to capture environmental statistics as well as subjective ...

This paper states necessary and sufficient conditions for the existence, uniqueness, and updating according to Bayes’ rule, of subjective probabilities representing individuals’ beliefs. The approach is preference based, and the result is an axiomatic subjective expected utility model of Bayesian decision making under uncertainty with statedependent preferences. The theory provides foundations ...

A Bayesian network is a powerful graphical model. It is advantageous for real-world data analysis and finding relations among variables. Knowledge presentation and rule generation, based on a Bayesian approach, have been studied and reported in many research papers across various fields. Since a Bayesian network has both causal and probabilistic semantics, it is regarded as an ideal representat...

Bandwidth plays an important role in determining the performance of nonparametric estimators, such as the local constant estimator. In this paper, we propose a Bayesian approach to bandwidth estimation for local constant estimators of time–varying coefficients in time series models. We establish a large sample theory for the proposed bandwidth estimator and Bayesian estimators of the unknown pa...