UMBC Home Page for Andrei Draganescu

Andrei Draganescu
Associate Professor, Department of Mathematics and Statistics, UMBC
Applied Mathematics Graduate Program Director

Ph. D., Applied Mathematics, University of Chicago, 2004
B.Sc., Mathematics, University of Bucharest, Romania, 1993

About me

I joined the Department of Mathematics and Statistics at the University of Maryland, Baltimore County as an assistant professor in Fall 2006 after completing a two-year postdoctoral appointment in the Optimization and Uncertainty Estimation Department at the Sandia National Laboratories in Albuquerque, New Mexico. I received my Ph.D. in applied mathematics from the University of Chicago in 2004, my dissertation being directed by Todd Dupont and Ridgway Scott. My research area is numerical analysis of partial differential equations, with current focus on multilevel algorithms for PDE-constrained optimization.

Teaching

Grants

Current research interests

Students

Selected publications

Contact

Links

Sayas Numerics Seminar

Teaching

Graduate student seminar schedule

Spring 2022 - Math 630: Numerical Linear Algebra, Mo We 5:30 - 6:45 pm, MP 120

Previously taught courses

Grants

2019 - 2022, NSF Award 1913201, Collaborative Research: Multilevel Methods for Optimal Control of Partial Differential Equations and Optimization-Based Domain Decomposition, PI with co-PI Bedrich Sousedik, in collaboration with Harbir Antil at George Mason University.

2014 - 2018, DOE Award DE-SC0005455 (renewal), Efficient Solution Methods for Large-Scale Optimization Problems Constrained by Time-Dependent Partial Differential Equations, PI.

2010 - 2013, DOE Award DE-SC0005455, Multilevel Techniques for Large-Scale Inverse Problems, PI.

2010 - 2013, NSF Award 1016177, Multilevel Methods in PDE Constrained Optimization, PI.

2008 - 2011, NSF Award 0821311, SCREMS: Parallel Computing for Interdisciplinary Research in Mathematics and Statistics, co-PI with Matthias Gobbert (PI), Nagaraj Neerchal, and Florian Potra.

2007, Sandia Corporation Contract, Mathematical Investigations in Support of Atomistic to Continuum Multiscale Coupling Analysis, PI, with Sandia sponsor Rich Lehoucq

Current research interests

Multilevel methods for PDE constrained optimization

Former postdocs

Ana Maria Soane (US Naval Academy)

Students

Michael Retzlaff

Sumaya Alzuhairy, Ph.D 2021

Mona Hajghassem, Ph.D. 2017 (Montgomery College)

Jyoti Saraswat, Ph.D. 2014 (Thomas More College)

Narae Lee (B.S. 2012)

Selected publications

Submitted

Sharp Estimates for the Convergence Rate of Orthomin(k) for a Class of Linear Systems (with Florin Spinu), submitted, arXiv:1109.2669.

Published or accepted

Optimal order multigrid preconditioners for the distributed control of parabolic equations with coarsening in space and time (with Mona Hajghassem), Optimization Methods and Software, DOI: 10.1080/10556788.2021.2022145, 2022 [link].
A Note on Multigrid Preconditioning for Fractional PDE-Constrained Optimization Problems (with Harbir Antil and Kiefer Green), Results in Applied Mathematics, Vol. 9, 2021, arXiv:2010.14600.
Algebraic multigrid preconditioning of the Hessian in PDE-constrained optimization (with Andrew Barker), Numerical Linear Algebra with Applications, Vol. 28 (1), 2021, arxiv:2001.11533.
Testing Homogeneity of Several Normal Population Means Based on an Interval Hypothesis (with Junyong Park), Communications in Statistics - Simulation and Computation, DOI: 10.1080/03610918.2019.1639737, 2019.
Multigrid preconditioners for the Newton-Krylov method in the optimal control of the stationary Navier-Stokes equations (with Ana Maria Soane), SIAM J. Numer. Anal., Vol. 57, No. 3, pp. 1494-1523, 2019. arXiv:1803.03890
Optimal order preconditioners for linear systems arising in the semismooth Newton solution of a class of control-constrained problems (with Jyoti Saraswat), SIAM. J. Matrix Anal. & Appl., 37(3), 1038-1070, 2016. Also available as arXiv:1411.6065.
Multigrid Preconditioning of Linear Systems for Semismooth Newton Methods Applied to Optimization Problems Constrained by Smoothing Operators, Optimization Methods and Software (DOI:10.1080/10556788.2013.854356), Vol. 29, No. 4, 786-818, 2014. Also available as arXiv:1106.0090.
Multigrid Solution of a Distributed Optimal Control Problem Constrained by the Stokes Equations (with Ana Maria Soane), Applied Mathematics and Computation, Volume 219, Issue 10, pp. 5622-5634, 2013. Also available as arXiv:1110.5281.
Multigrid Preconditioning of Linear Systems for Interior Point Methods Applied to a Class of Box-constrained Optimal Control Problems (with Cosmin Petra), SIAM Journal on Numerical Analysis, Vol. 50, No. 1, pp. 328-353, 2012.
Optimal Order Multilevel Preconditioners for Regularized Ill-posed Problems, (with Todd Dupont), Mathematics of Computation, 77, no. 264, pp. 2001-2038, 2008.
Dynamic Data-Driven Inversion for Terascale Simulations: Real-Time Identification of Airborne Contaminants, (with Volkan Akcelik, George Biros, Omar Ghattas, Judith C. Hill, and Bart van Bloemen Waanders), Proceedings of SC2005 (2005), IEEE/ACM.
Failure of the discrete maximum principle for an elliptic finite element problem, (with Todd Dupont and L. Ridgway Scott), Mathematics of Computation, 74, no. 249, pp. 1-23, 2005.

Tech Reports

Hamiltonian Molecular Dynamics for Computational Mechanicians and Numerical Analysts (with Rich Lehoucq and Paul Tupper), Sandia National Laboratories, Tech. Report, SAND 2008-6512, April 2009.
A Fast Multigrid Method for Inverting Linear Parabolic Problems, University of Chicago, Department of Computer Science, Tech. Report, TR-2004-01, January 2004.

Past Conferences

UMBC Finite Element Circus, October 20-21, 2017

Contact

University of Maryland, Baltimore County
Department of Mathematics and Statistics
1000 Hilltop Circle
Baltimore, MD 21250, USA

office: MP 420
phone: (410) 455-3237
fax: (410) 455-1066
draga 'at' umbc.edu
userpages.umbc.edu/~draga