Multi-Agent Bayesian Framework For Parametric Selection In The Detection And Diagnosis of Tuberculosis Contagion In Nigeria

Abstract

Decision making has become quite a critical factor in our everyday living. The provision of data alongside its consequent processing has further sought to extend and expand our reasoning faculties as well as effectively aid proper decision making. But data is daily, produced at an exponential rapid rate and the volume in amount of data churned out to be processed even more so that we now require data storage optimization techniques to process such humongous volume of data. These have today, necessitated the need for advancement in data mining process. With the tremendous advances made in data mining, machine learning, storage virtualization and optimization – amongst other fields of computing – researchers now seek a new paradigm and platform called data science. This field today has become quite imperative as it seeks to provide beneficial support in constructing models and algorithms that can effectively assist domain experts and practitioners to make comprehensive and sound decisions regarding potential problematic cases. We focus on modeling social graph using implicit suggest algorithm in medical diagnosis to effectively respond to problematic cases of Tuberculosis (TB) in Nigeria. We introduce spectral clustering and Bayesian Network, construct algorithms cum models for predicting potential problematic cases in Tuberculosis as well as compare the algorithms based on data samples collected from an Epidemiology laboratory at the Federal Medical Center Asaba in Delta State of Nigeria. The volume of data was collated and divided into two data sets which are the training dataset and the investigation dataset. The model constructed by this study has shown a high predictive capability strength compared to other models presented on similar studies.