Vahan Hovhannisyan, Ph.D. student.
My research interests are in development and analysis of optimisation algorithms for large-scale problems, particularly arising from statistical machine learning and computer vision applications.
MAGMA: Multi-level Accelerated Gradient Mirror Descent Algorithm for Large Scale Convex Composite Minimization, with Panos Parpas and Stefanos Zafeiriou, SIAM Journal on Imaging Sciences, BibTex.
“Multi-Level Accelerated Algorithm for Large-Scale Convex Composite Minimization”. SIAM Conference on Imaging Science (invited presentation), Albuquerque, NM, 2016.
“An accelerated multiresolution algorithm for large scale non-smooth convex optimisation”, 17th British-French-German Conference on Optimization, London, 2015.
Magma - a Matlab framework of optimisation algorithms with applications in machine learning.
Visiting lecturer Imperial College Business School for Optional C++ course
Guest teacher at London School of Economics and Political Science for
Tutor at Imperial College London for
Graduate teaching assistant at Imperial College London for
Business Analytics, Computational Finance, Software Engineering: Algorithms, Operations Research.
Master of Science in Applied Mathematics from ETH Zurich Switzerland (2013)
Bachelor of Science in Mathematics (with distinction) from State Engineering University of Armenia (2009)
Address: Department of Computing, Imperial College, 180 Queen's Gate, London, SW7 2AZ
Office: Room 302, Huxley Building
Phone: +44 (0) 20 75948374
Email: vh13 AT imperial.ac.uk