Recent Developments on Probabilistic Graphical Model Applied in Data Analysis

Probabilistic graphical models (PGMs) are recognized to strongly trapping the dynamics of physical systems such as data-driven, communication, imaging, security, and allied fields. These models are using to characterize mutually the physical properties of a distributed complex system. This has led to discovery of several PGM techniques of interests include Bayesian networks (BN), Gaussian graphical models (GGMs), graphical Markov models (GMMs) which are the straightforward components that establish the most sophisticated intuitive diagrams of complex conditional dependencies between a set of random variables as well as relationships between stochastic variables. Thereby, recent achievements in employing the graphical models, by purchasing new algorithms and concepts have been proposed to derive and enhance the efficiency of such models. In this perspective, this paper presents the latest progress of graphical models which has received considerable attention in broadband literature. It provides a rapid underlying and understanding of the graphical model and its potential applications with new insight into the massive interests involved in computer and information technologies. Therefore, this can be serves as pivotal guide that illustrate update information such as research direction, suitability, and limitation of the graphical model recently occur in computer science & information technology. So far, this review is attributed to providing a robust background to both industry and science community on ongoing fascinate applications in modern information technology. In addition, a simplified illustration of Bayesian and Markov approaches are depicted in Annex A and Annex B respectively.