Hidden Markov Model of Disease Progression and Control with Reference to COVID-19 Spread

Abstract

Disease progression studies through stochastic modeling are the most effective approaches as different processes involved in the disease acquisition, growth, spread, and control are random. This study develops a stochastic model for studying the disease spread using Markov Processes (MP) and Hidden Markov Models (HMM). This study considered two states of illness under the categories of hidden and visible. Further hidden states, as well as visible states, are classiffed into two groups each. This study attempted to relate the spread of disease in Tamil Nadu and Puducherry and its neighboring states. Increment/Decrement in daily positive cases of Tamil Nadu and Puducherry in uence the Increment/ Decrement in neighboring states' daily positive cases, assuming there are regular transitions of patients from one place to another. This study develops HMM for transitions among different states (Increment/Decrement) for understanding the dynamics of positivity for two consecutive days and three days. Probability distributions of the prevalence of positivity are derived from the developed transition probability matrices. The study further derived different statistical measures mathematical/ functional relations through the parameters under consideration. This study will help to measure the severity of the disease spread. The development of an interactive user interface for healthcare management will be the scope of this study.