Stochastic filtering based transmissibility estimation of novel coronavirus

Area of Science:

• Epidemiology
• Mathematical Biology
• Control Theory

Background:

• Accurate estimation of COVID-19 transmissibility is crucial for effective public health interventions.
• Traditional epidemiological models may not fully capture the complex dynamics of novel infectious diseases.
• Advanced filtering and signal processing techniques can enhance the accuracy of disease transmission modeling.

Purpose of the Study:

• To develop and evaluate novel methods for estimating COVID-19 transmissibility.
• To apply generalized fractional-order calculus (FOC) based extended Kalman filter (EKF) and wavelet transform (WT) to the bats-hosts-reservoir-people (BHRP) model.
• To compare the performance of the proposed models against actual COVID-19 data from India and China.

Main Methods:

• State-space representation of the susceptible-exposed-infectious-recovered (SEIR) model using fractional order differential equations.
• Application of Extended Kalman Filter (EKF) to the discrete vector representation of the BHRP model, accounting for process and measurement noise.
• Utilizing Kronecker product based Wavelet Transform (WT) methods for transmissibility estimation.

Main Results:

• The proposed FOC-based EKF and WT models provide accurate estimations of COVID-19 transmissibility.
• The models demonstrate superior performance by considering both process and measurement noise, leading to more probable states.
• Comparisons with actual data from India and China show good agreement for estimated infectious and recovered populations.

Conclusions:

• Generalized fractional-order calculus combined with EKF and WT offers a robust framework for estimating infectious disease transmissibility.
• The proposed models are versatile and encompass conventional EKF and WT methods as special cases.
• These advanced modeling techniques can significantly improve the understanding and management of pandemics like COVID-19.