Objectives: To develop a reliable mathematical model in order to predict the evolution of various epidemiological factors and parameters for COVID-19 across the globe. Methods: A novel dynamic Susceptible-Exposed-Infected-Recovered-Died (SEIRD) model is proposed in this research. The proposed, two-step approach assumes the infection rate which is dependent on time, to estimate the evolution of various variables of the model. In the first step, parameters like clinical and transmission are estimated, whereas in the second step, simulation of the model is done to predict the outbreak. Findings: Making use of this model, the total number of people who are likely to be afflicted by an infectious disease in a closed population over a period of time can be computed theoretically. Novelty: The proposed model results into low computational complexity since it is deterministic in nature. Secondly, SEIRD model equations are solved in frequency domain that converts the integraldifferential equations into simple algebraic equations. This further reduces the computational burden.