Jose Luis Morales

Jose Luis Morales
morales -at- ece.northwestern.edu

Education:
- B. Sc. degree in Chemistry, UNAM, Mexico.
- B. Sc. degree in Mathematics, UNAM, Mexico.
- M. Sc. degree in Computer Science, UNAM, Mexico.
- Ph. D. Imperial College of Science Technology and Medicine, UK.
Position:
- Professor. Instituto Tecnologico Autonomo de Mexico (ITAM). Departamento de Matematicas.
Projects:
- Google project. Data in compressed format.
- Interior-point methods for large-scale nonlinear programming. In collaboration with The Optimization Technology Center Team.
- Development and implementation of preconditioning techniques. In collaboration with The Optimization Technology Center Team.
Recent articles and technical reports:
1. Automatic Preconditioning by Limited Memory Quasi-Newton Updates, (2000), SIAM, Journal on Optimization, 10, 4, pp. 1079-1096, with Jorge Nocedal.
2. Algorithm 809: PREQN: Fortran 77 for Preconditioning the Conjugate Gradient Method. (2001), ACM Transactions on Mathematical Software, 27, 1, pp. 83-91, with Jorge Nocedal.
3. Enriched Methods for Large-Scale Unconstrained Optimization, (2002), Computational Optimization and Applications (COAP), 21, 2 pp. 143-154, with Jorge Nocedal.
4. A Numerical Study of Limited Memory BFGS Methods,(2002), Applied Mathematics Letters, 15, 4, pp. 481-487.
5. Assessing the Potential of Interior Methods for Nonlinear Optimization. (2003). Large-Scale PDE-Constrained Optimization. L.T. Biegler, O. Ghattas, M. Heinkenschloss, and B. van Bloemen Wanders (Eds.) Series: Lecture Notes in Computational Science and Engineering. Volume. 30, pp 167-183. Springer-Verlag. Heidelberg, with Jorge Nocedal, Richard Waltz, Guanghui Liu and Jean-Pierre Goux.
  In this paper we report a series of numerical experiments with interior point (LOQO, KNITRO) and active-set SQP codes (SNOPT, filterSQP). The tests were performed on a set of small, medium-size, and moderately large problems. We performed the experiments using a set of 509 problems from the CUTE collection that have been translated into AMPL by Benson and Vanderbei. Since the sample is quite large, we report our results using the performance profiles proposed by Dolan and Moré.
  Detailed outputs (revised versions: November 13, 2001), for the complete Benson-Vanderbei 740 problems sample, are available:
  LOQO
  filterSQP
  KNITRO
  SNOPT
6. An Interior Algorithm for Nonlinear Optimization that Combines Line Search and Trust Region Steps (2006): Mathematical Programming Series A, 107, 3, pp. 391--408. Published Online First, 25 November 2005, with Richard Waltz, Jorge Nocedal and Dominique Orban.
Software:
- Software for preconditioning the Conjugate Gradient method, in collaboration with Jorge Nocedal. The package PREQN is freely available and may be used for research, educational or comercial purposes. However, we expect that all research publications using the package give reference to the papers describing our software.