Author: Suman K. Rao
Abstract: Bayesian methods and inference provide a powerful probabilistic framework for learning from data under uncertainty. Unlike classical frequentist approaches, Bayesian inference treats unknown parameters as random variables and incorporates prior knowledge through probability distributions. Over the last few decades, Bayesian methodology has seen significant development driven by advances in computation, particularly Markov Chain Monte Carlo and variational inference techniques. These developments have enabled Bayesian methods to be applied in complex high-dimensional problems across diverse domains such as machine learning, engineering, economics, medical diagnostics, and environmental modeling. This paper presents a comprehensive review of Bayesian inference, beginning with foundational concepts such as Bayes’ theorem and prior– posterior updating. Computational strategies for posterior estimation are discussed in detail, followed by an overview of hierarchical models, model comparison, and Bayesian decision theory. Practical applications and emerging trends are also highlighted. The paper aims to serve as a self-contained reference for researchers and postgraduate students in engineering mathematics and statistics.
Keywords: Bayesian inference, prior distribution, posterior distribution, Markov Chain Monte Carlo, probabilistic modeling, uncertainty quantification
Full Issue
| View or download the full issue | PDF 115-125 |