Patricia L. Gillett-Kawamoto

Researcher and Developer

Your Name

Hello!

I'm a trained computational mathematician and operations researcher. My passions are for scientific computing and the modeling and optimizing of systems. I love being part of the implementation, especially the juicy mathematically intense bits.

Since 2016 I've been building recommender systems with the Discovery Algorithms team at Shopify and it's been awesome. Drop me a line or learn more about openings in Shopify data!

Experience

Sept 2016 - present

Shopify, Data Developer

Designing and building data products, particularly recommender systems, to inject intelligence into various aspects of the Shopify product.

Sept 2010 - Aug 2016

Ecole Polytechnique Montreal, Research Assistant (PhD student)

Nonlinear optimization project:
Developing strong SDP relaxations of QPCCs (quadratic problems with complementarity constraints) and leveraging them to develop better local and global solution methods for this class of problems.

Software project ('PyQPCC'):
Part mathematical optimization, part scientific computing, and part glue code and interfaces, this suite of tools enables a user to build a Python-based methodology for QPCCs and/or their SDP relaxations while having access to solvers and test problem generators which are native to Matlab, AMPL, and GAMS. The project's open source release is pending.

Sept 2010 - Aug 2016

Student member of GERAD

Education

PhD candidate, Engineering Mathematics

2010-present, Polytechnique Montreal

B. Math Honours in Computational Mathematics, Economics option

2005-2010, University of Waterloo

Undergraduate exchange student, Department of Informatics

2008-2009, Kyoto University, Japan

Diploma in Japanese Language

2006-2010, Renison University College

Publications

J. Kawamoto, P. L. Gillett, “Frequency-based Constraint Relaxation for Private Query Processing in Cloud Databases,” Proc. of the 27th Annual IEEE Canadian Conference on Electrical and Computer Engineering, pp.1275-1280, Toronto, May 2014.

J. Kawamoto, P. L. Gillett, J. Sakuma, プライベート問合せにおける問合せ頻度を用いた制約緩和手法 (Frequency-based Constraint Relaxation for Private Query), IPSJ Transactions on Databases, Vol.6, No.3, pp.50-60, June 2013 (in Japanese).

Presentations

P. L. Gillett, M. F. Anjos, J. Judice, Combining semidefinite relaxations and NLP solvers for improved feasible solutions of QPLCCs, INFORMS, Philadelphia, Nov 2015

P. L. Gillett, M. F. Anjos, J. Judice, Combining semidefinite relaxations and NLP solvers for improved feasible solutions of QPLCCs, CORS-INFORMS, Montreal, June 2015

P. L. Gillett, M. F. Anjos, Finding better solutions to nonconvex quadratic equilibrium problems using semidefinite programming, Optimization Days, Montreal, May 2014.

P. L. Gillett, M. F. Anjos, Semidefinite programming approaches for a class of complementarity problems, Optimization Days, Montreal, May 2012.

P. L. Gillett, M. F. Anjos, Semidefinite programming approaches for a class of complementarity problems, Meet a GERAD Researcher Seminar Series, Montreal, March 2012.

P. L. Gillett, M. F. Anjos, Semidefinite programming approaches for complementarity problems, Optimization Days, Montreal, May 2011.

Poster Presentations

P. L. Gillett, M. F. Anjos, A semidefinite programming approach for nonconvex quadratic optimization with complementarity constraints, Women Optimize in the West, Calgary, June 2013.

P. L. Gillett, M. F. Anjos, A semidefinite programming approach for nonconvex quadratic Optimization with complementarity constraints, Polynomial Optimisation Workshop, Cambridge, UK, July 2013.

Skills

Languages and packages

Scientific computing in Python:

  • General (numpy, scipy)
  • Convex optimization (cvxpy, cvxopt)
  • Using the NEOS server (xml-rpc)
  • Data analysis (pandas)
  • Data visualization (matplotlib, seaborn)
  • Machine learning (scikit-learn)

Mathematical languages:

Other languages:

Languages I have history with but
have not used recently:

C, C++, Java, Scheme

Working methods:

Let's keep in touch!

E-mail (Work): trish.gillett@shopify.com
E-mail (Personal): discardthree@gmail.com
Located in Montreal, QC, Canada.