Fractional and fractal extensions of epidemiological models
Authors:
Enrique C. Gabrick,
Ervin K. Lenzi,
Antonio M. Batista
Abstract:
One way to study the spread of disease is through mathematical models. The most successful models compartmentalize the host population according to their infectious stage, e.g., susceptible (S), infected (I), exposed (E), and recovered (R). The composition of these compartments leads to the SI, SIS, SIR, and SEIR models. In this Chapter, we present and compare three formulations of SI, SIS, SIR, a…
▽ More
One way to study the spread of disease is through mathematical models. The most successful models compartmentalize the host population according to their infectious stage, e.g., susceptible (S), infected (I), exposed (E), and recovered (R). The composition of these compartments leads to the SI, SIS, SIR, and SEIR models. In this Chapter, we present and compare three formulations of SI, SIS, SIR, and SEIR models in the framework of standard (integer operators), fractional (Caputo sense), and fractal derivatives (Hausdorff sense). As an application of the SI model, we study the evolution of AIDS cases in Bangladesh from 2001 to 2021. For this case, our simulations suggest that fractal formulation describes the data well. For the SIS model, we consider syphilis data from Brazil from 2006 to 2017. In this case, the three frameworks describe the data with good accuracy. We used data from Influenza A to adjust the SIR model in previous approaches and observed that the fractional formulation was better. The last application considers the COVID-19 data from India in the range 2020-04-10 to 2020-12-31 to adjust the parameters of the SEIR model. The standard formulation fits the data better than the other approaches. As a common result, all models exhibit steady solutions in the different formulations. The time to reach a steady solution is correlated to the considered approach. The standard and fractal formulations reach the steady state earlier when compared with the fractional formulation.
△ Less
Submitted 7 September, 2024;
originally announced September 2024.
Fractional dynamics and recurrence analysis in cancer model
Authors:
Enrique C. Gabrick,
Matheus R. Sales,
Elaheh Sayari,
José Trobia,
Ervin K. Lenzi,
Fernando da S. Borges,
José D. Szezech Jr.,
Kelly C. Iarosz,
Ricardo L. Viana,
Iberê L. Caldas,
Antonio M. Batista
Abstract:
In this work, we analyze the effects of fractional derivatives in the chaotic dynamics of a cancer model. We begin by studying the dynamics of a standard model, {\it i.e.}, with integer derivatives. We study the dynamical behavior by means of the bifurcation diagram, Lyapunov exponents, and recurrence quantification analysis (RQA), such as the recurrence rate (RR), the determinism (DET), and the r…
▽ More
In this work, we analyze the effects of fractional derivatives in the chaotic dynamics of a cancer model. We begin by studying the dynamics of a standard model, {\it i.e.}, with integer derivatives. We study the dynamical behavior by means of the bifurcation diagram, Lyapunov exponents, and recurrence quantification analysis (RQA), such as the recurrence rate (RR), the determinism (DET), and the recurrence time entropy (RTE). We find a high correlation coefficient between the Lyapunov exponents and RTE. Our simulations suggest that the tumor growth parameter ($ρ_1$) is associated with a chaotic regime. Our results suggest a high correlation between the largest Lyapunov exponents and RTE. After understanding the dynamics of the model in the standard formulation, we extend our results by considering fractional operators. We fix the parameters in the chaotic regime and investigate the effects of the fractional order. We demonstrate how fractional dynamics can be properly characterized using RQA measures, which offer the advantage of not requiring knowledge of the fractional Jacobian matrix. We find that the chaotic motion is suppressed as $α$ decreases, and the system becomes periodic for $α\lessapprox 0.9966$. We observe limit cycles for $α\in (0.9966,0.899)$ and fixed points for $α<0.899$. The fixed point is determined analytically for the considered parameters. Finally, we discover that these dynamics are separated by an exponential relationship between $α$ and $ρ_1$. Also, the transition depends on a supper transient which obeys the same relationship.
△ Less
Submitted 25 August, 2023;
originally announced September 2023.
