Manu Jayadharan

Advisor: Ivan Yotov

Research Interests: Scientific Machine learning, Numerical analysis, CFD, Finite element methods, Math Finance, Domain Decomposition and HPC.

Quick links :

Github: https://github.com/mjayadharan
Linkedin: https://www.linkedin.com/in/manu-jayadharan/
Research gate: https://www.researchgate.net/profile/Manu_Jayadharan
FluidLearn: https://pypi.org/project/fluidlearn/

Recent Publication list: Domain decomposition and partitioning methods for mixed finite element discretizations of the Biot system of poroelasticity. 

                                      Link https://arxiv.org/abs/2010.15353

About 

I am a computational scientist with expertise in numerical analysis and scientific machine learning. Plese visit my github profile to see the latest projects that I work on.

I have experience in developing and implementing novel algorithms to solve multiphysics CFD problems using both data-driven deep learning techniques and classical mixed finite element methods (FEM). I have created and currently maintains many open-source software packages to solve complex system of partial differential equations (PDEs). I am also interested in application of ML techniques to financial and risk models.

My current area of research lies at the intersection of machine learning and numerical analysis, commonly called scientific machine learning. Scientific ML is a very recent and emerging field of applied mathematics where we develop techniques to efficiently solve complex partial differential equations, using the latest advances in deep learning, combined with well-established classical techniques, like finite element methods.

My most recent project include developing an open-source python package based on TensorFlow2, named FluidLearn. This package implements physics informed neural networks (PINNs), which is intended to make use of artificial neural networks (ANNs) to solve any well-posed non-linear PDE defined on complex domains. A pre-production release of the package is available on GitHub and PyPi.

I have prior experience working in diverse fields like computational biology, stochastic process modelling and mathematical analysis. I have experience working with a variety of programming languages including C++, perl, python, matlab and fortran, though I primarily work with python and C++ now.