Favorite links
U. Waterloo
Workshops and conferences

SIAM Conference on
Optimization, San Diego, 2014 May 1922

Householder Symposium
(numerical linear algebra), Spa, Belgium, 2014 June 813

Workshop
on optimization and matrix methods in big data, part of the Thematic Program
in Big Data, Fields Institute, Toronto, Canada, 2015 February 913.

International Symposium on Mathematical
Programming, Pittsburgh, PA, 2015 July 1217.

SIAM Conference
on Optimization, Vancouver, B.C., 2017 May 2225, featuring a
minisymposium on "Recent Progress in FirstOrder Methods for Convex
Optimization".

Householder Symposium
on Numerical Linear Algebra, Blackburg, VA, 2017 June 1823.

MOPTA (Modeling and Optimization:
Theory and Applications), Bethlehem, PA, 2017 August 1618.

International Symposium on Mathematical
Programming, Bordeaux, France, 2018 July 17 featuring an
8talk
minisymposium on "Firstorder methods for largescale convex
optimization."

Statistical Inference, Learning and
Models in Data Science, Fields Institute, Toronto, 2018 Sep 2428.

Workshop on lowrank models and applications, Mons, Belgium, 2019 Sep 1213.
Professional societies:
Research supported by:

Associate Dean, Computing, Faculty of Mathematics
Director, Mathematics Faculty Computing Facility
(MFCF)
Professor (PhD, Stanford, 1989),
Department of Combinatorics & Optimization
MC 6304 (C&O office)
MC 3047 (MFCF office)
University of Waterloo
200 University Avenue W.
Waterloo, ON N2L 3G1
Canada
On Github, I am user StephenVavasis. I am an inactive user of Facebook.
Anyone claiming to be me on LinkedIn or other social networks is
probably an impostor.
email: vavasis@uwaterloo.ca
phone: +15198884567 ext. 32130 or 37719
fax: +15197255441
Courses
In Fall 2018, I taught CO673/CS794, Optimization for Data Science.
This course has its website on
UWaterloo Learn.
For those without Learn access yet, you can access
courseinformation material
and Problem Set 1 here.
Research and Publications
My research interests are
continuous
optimization and
data science.
More specifically,
I am interested in:

Efficient algorithms for optimization,

Applications of optimization to data science,

Numerical linear algebra in optimization and data science

Complexity issues in continuous optimization.
I have also in the past been interested in
geometric problems arising in scientific computing
and numerical linear algebra arising in differential equations.
Some publications and recent manuscripts:

S. Vavasis and Y. Ye, ``A primal dual accelerated interior
point method whose running time depends only on A''
(compressed postscript, 170k).
Final version appeared in Math. Progr. 74 (1996) 79120.

P. Hough and S. Vavasis, ``Complete orthogonal decomposition
for weighted least squares''
(gzip'd
postscript, 84k;
plain
postscript, 330k). Final version appeared in SIAM J. Matrix Anal. Appl.,
18 (1997) 369392.

S. Mitchell
and S. Vavasis, ``An aspect ratio bound for triangulating a dgrid
cut by a hyperplane''
(compressed
postscript, 118k;
plain
postscript, 339k). Extended abstract appeared in Proc. 1996
Symp. Comput. Geom., ACM Press, 4857.
 T.
Driscoll and S. Vavasis,
``Numerical conformal mapping using crossratios and Delaunay triangulation''
(gzip'd postscript, 139k;
plain
postscript, 395k).
Final version appeared in
SIAM J. Sci. Comp. 19 (1998) 17831803.

S. Mitchell
and S. Vavasis, ``Quality Mesh Generation in Higher Dimensions''
(gzip'd
postscript, 130k;
plain
postscript, 388k). Final version appeared in SIAM J. Comput.
29 (2000) 13341370.

E. Bobrovnikova and S. Vavasis,
``Accurate Solution of Weighted Least Squares by Iterative Methods''
(gzip'd postscript, 105k;
plain
postscript, 349k). Final version appeared SIAM J. Matrix An. App
22 (2001) 11531174.

E. Bobrovnikova and S. Vavasis,
``A Norm Bound for Projections with Complex Weights''
(gzip'd
postscript, 106k;
plain
postscript, 215k). Final version appeared Linear Alg. Appl.
307 (2000) 6975.

V. Howle
and S. Vavasis,
``Preconditioning complexsymmetric layered systems arising
in electric power modeling''
(gzip'd
postscript, 42k;
plain
postscript, 156k).

S. Vavasis, ``A note on efficient computation of the gradient in semidefinite
programming''
(gzip'd
postscript, 99k;
plain
postscript, 196k).

G. Jónsson
and S. Vavasis, ``Solving polynomials with small leading
coefficients''
(gzip'd
postscript, 100k;
plain
postscript, 290k). Final version appeared: SIAM J. Matrix Analysis App., 26 (2005) 400414.

V. Howle and S.
Vavasis, ``An iterative method for solving complexsymmetric systems arising
in electric power modeling''
(gzip'd
postscript, 123k;
plain
postscript, 343k). Final version appeared:
SIAM J. Matrix Analysis App.,
26 (2005) 11501178

G. Jónsson
and S. Vavasis, ``Accurate solution of polynomial equations using Macaulay
resultant matrices''
(gzip'd
postscript, 284k;
plain
postscript, 689k). Final version appeared:
Mathematics of Computation,
74 (2005) 221262

S. Vavasis, ``A BernsteinBezier sufficient condition for invertibility
of polynomial mapping functions'' (revised 3 Nov 2001)
(gzip'd
postscript, 100k;
plain
postscript, 200k).

K.
Papoulia and S. Vavasis, ``Time continuity in cohesive finite element
modeling'' (pdf, 319k).
Final version appeared:
International Journal for Numerical Methods in Engineering
58(5) 679701, 2003.

P. Ganguly, S. Vavasis and K.
Papoulia,
``An algorithm for twodimensional mesh generation based on the
pinwheel tiling''
(pdf,
289k) SIAM J. Scientific Comput., 28 (2006) 15331562.

E. Boman,
B. Hendrickson,
and S. Vavasis,
``Solving Elliptic Finite Element Systems in NearLinear Time
with Support Preconditioners''
(Arxiv link),
SIAM J. Numer. Anal. 46 (2008) 32643284.

K.
Papoulia,
S. Vavasis and P. Ganguly, Spatial convergence of crack nucleation
using a cohesive finite element model on a pinwheelbased mesh,
International J. Numer. Meth. Eng. (pdf, 291k), 67(2006) 116.

ChinHang Sam,
K.
Papoulia,
Stephen Vavasis, Obtaining initially rigid cohesive finite element models that are temporally convergent,
Engineering Fracture Mechanics, 72 (2005) 22472267.
(pdf, 310k)

S. Shontz and
S. Vavasis,
``A linear weighted laplacian smoothing framework for warping tetrahedral meshes''
(arxiv link)
Analysis of and workarounds for element reversal for a finite
elementbased algorithm for warping triangular and tetrahedral meshes,
BIT Numerical Mathematics, Volume 50 (2010), Issue 4, 863884.

Gun Srijuntongsiri and S. Vavasis,
A Fully Sparse Implementation of a PrimalDual InteriorPoint Potential Reduction Method for Semidefinite Programming,
submitted to SIAM Journal on Optimization
(arxiv link)

S. Vavasis,
A conjecture that the roots of a univariate polynomial lie in a union
of annuli
(arxiv link)

Gun Srijuntongsiri and
S. Vavasis,
A condition number analysis of a linesurface intersection algorithm
(arxiv link).
SIAM J. Sci. Comput 30 (2008) 10641081.

S. Shontz and
S. Vavasis,
``A robust solution procedure for hyperelastic solids with
large boundary deformation''
(arxiv link).
Final version appeared in Engineering with COmputers, posted on
line 2011June11, print edition to appear.

Gun Srijuntongsiri and S. Vavasis,
Properties of polynomial bases used in linesurface
intersection algorithm
(arxiv link)

S. Vavasis,
On the complexity of nonnegative matrix factorization
(arxiv link).
Final version appeared, SIAM J. Optim. Volume 20, Issue 3, pp. 13641377 (2009).

Gun Srijuntongsiri and
S. Vavasis,
A condition number analysis of a surfacesurface intersection algorithm
(arxiv link)

M. Biggs, A. Ghodsi and S. Vavasis, Nonnegative matrix factorization
via rankone downdate,
preliminary journal version from
arxiv.org, ICML 2008
conference
version.

B. Ames and S. Vavasis,
Nuclear norm minimization for the planted clique and biclique problems
(arxiv link;
published paper.)

B. Ames and S. Vavasis,
Convex relaxation for the planted kdisjointclique problem
(arxiv link)

X.
V. Doan and S. Vavasis,
Finding approximately rankone submatrices with the nuclear
norm and l_{1} norm
(arxiv link).
Unpublished proof of NPhardness of LAROS problem.

X.
V. Doan,
K.C. Toh and S. Vavasis,
A proximal point algorithm for sequential feature
extraction applications
(arxiv link)

S. Karimi and S. Vavasis,
Detecting and correcting loss of independence in nonlinear
conjugate gradient (arxiv link)
(previously "Conjugate gradient with subspace
optimization").

N. Gillis
and S. Vavasis,
Fast and Robust Recursive Algorithms for Separable Nonnegative Matrix Factorization (arxiv link).

S. Sastry, S. Shontz, and S. Vavasis, A logbarrier method for mesh quality
improvement and untangling, Final version appeared in Engineering with Computers,
Nov 2012 (Springer Link web publication). Preliminary version appeared in
Proceedings of the 2011 International Meshing Roundtable.

L. Elkin,
Ting Kei Pong and S. Vavasis,
Convex relaxation for finding planted influential nodes
in a social network
(arxiv link).

S. Vavasis, Some notes on applying computational divided differencing
in optimization
(arxiv link).

N. Gillis
and S. Vavasis
Semidefinite programming based preconditioning
for more robust nearseparable nonnegative matrix factorization,
SIAM J. Optimiz., 25 (2015) 677698,
(arxiv link).

D. Drusvyatskiy,
S.Vavasis, H. Wolkowicz
Extreme point inequalities and geometry of the rank sparsity ball,
Math. Progr. A 152 (2015) 521544,
(arxiv link).

X.
V. Doan and S. Vavasis,
Finding the largest lowrank clusters
with Ky Fan 2knorm and
l_{1} norm
(arxiv link).

S. Karimi &
S. Vavasis,
IMRO: a proximal quasiNewton method for
solving l1regularized least squares problem,
(arxiv link).

N. Gillis &
S. Vavasis,
On the complexity of robust PCA and
l_{1}norm lowrank matrix approximation,
arxiv link).

S. Karimi &
S. Vavasis,
A unified convergence bound for conjugate gradient and accelerated
gradient,
(arxiv link).

S. Karimi &
S. Vavasis,
A single potential governing convergence of conjugate gradient, accelerated gradient and geometric descent,
(arxiv link).

S. Vavasis, K. Papoulia and
M. Hirmand,
A secondorder cone interiorpoint method for initially rigid
cohesive fracture,
PDF.

T. Jiang, S. Vavasis and C. W. Zhai,
Recovery of a mixture of Gaussians
by sumofnorms clustering.

C. Paquette and S. Vavasis,
Potentialbased analyses of firstorder methods for constrained and composite
optimization
