I am a Ph.D. candidate in Computational Science at Middle Tennessee State University. My advisor is Dr. Abdul Khaliq. My research interest is on Data-driven deep learning algorithms for differential equation models. My codes will be on github at https://github.com/okayode. My work involves finding efficient numerical Solutions of partial differential equations and systems, parameter estimation and inverse problems. Presently, I am working on epidemiological models for an infectious disease such as COVID-19, with data from the publicly accessible COVID-19 data. We proposed that mitigations such as governmental actions and public responses, forces the model parameters to become time series. In our approach, we do not need to explicitly specify the time-dependent model parameters, instead we learn these parameters by defining them to be outputs of multilayer perceptrons (MLPs) that are connected to a larger MLP that predicts the solution to the epidemiological model. We consider when there is a spatial structure in the model and when there is none.

I am also interested in the discovery of governing equations of reaction-diffusion models such as the FitzHugh-Nagumo equations with nonlocal diffusion. In these models, the data-driven approach is able to detect nonlinear diffusion term from sparse experimental data.

Email: | kayode.olumoyin@mtsu.edu |

Office: | Davis Science Building 148 |

Phone: | 419 315 5667 |