|
Michael C. Grant, Ph.D. d/b/a CVX Research 1104 Claire Ave. Austin, TX 78703-2502 (512) 647-2097 mcg@cvxr.com |
Consultancy
In addition to support for CVX and TFOCS, I am available for short-term and part-time contracts, including the following:
- Application-related modeling assistance
- Training and short courses
- Custom algorithm design and implementation
Areas of expertise include:
- Convex optimization
- Compressive sampling
- Machine learning and statistical regression
- Fast, efficient computational algorithms
- Acceleration through exploitation of problem structure
- CPU, GPU, DSP, and ASIC implementations
- Applications: engineering design, scientific, financial
CVX: Matlab Software for Disciplined Convex Programming
CVX is a popular modeling framework for disciplined convex programming. CVX turns Matlab into a modeling language, allowing constraints and objectives to be specified using standard Matlab expression syntax. The result is a powerful tool for the rapid prototyping of models and algorithms incorporating convex optimization.
TFOCS: Templates for First-Order Conic Solvers
TFOCS (pronounced tee-fox) is a library designed to facilitate the construction of first-order methods for a variety of convex optimization problems. It was developed originally to support our study of sparse recovery models in this paper, but in fact it is applicable to a wider variety of models. TFOCS can scale to much larger model sizes than CVX. On the other hand, it requires all models to be manually converted to one of its standard forms; therefore, it requires more expertise to use.
Commercial support and service
As with many free open-source offerings, support for CVX and TFOCS is limited to bug fixes and simple improvements as time permits. CVX is licensed under the GPL, but alternative licensing arrangements may be negotiated. TFOCS requires a commercial license from CalTech for any commercial use; no-charge evaluation licenses are available. Feel free to contact me to address these matters and for additional support options, including expedited fixes, exclusive and/or expedited enhancements, and porting models to standalone applications.
Current projects and affiliations
Guest lecturer, The University of Texas at Austin
Department of Mechanical Engineering
Operations Research and Industrial Engineering
Course: Nonlinear Programming (focus: convex optimization)
Staff Scientist, California Institute of Technology
Department of Applied and Computational Mathematics
Division of Engineering and Applied Science
Advisor: Emmanuel Candès
Project: GPU acceleration of compressed sensing for wireless applications
Other projects subject to NDA.