The following are some of my research interests, reflected in selected downloadable writings.

• Network Optimization. This research focuses on methods and software for optimizing many varieties of network problems. Recent projects include:
   
• Interval flow networks: a new class of network models that restrict the flow on an arc to be either 0 or between a given lower and upper bound. Extremely efficient optimization and heuristic algorithms have been developed and tested.
• Network methodologies: parallel network algorithms, reoptimization techniques, and large-scale problems.
• Network applications: include telecommunications network design, railroad/airline logistics operations, optimal file matching, and targeting and strike planning for military air operations.
• Telecommunications: network design and planning tools. A Telecommunications Network Research Laboratory to support this work has been funded by the National Science Foundation Of special interest are algorithms for optimizing DWDM and all-optical broadband network designs (work sponsored by the Texas Advanced Technology Program).
   
• Computational Data Analytics.
   
• Optimization-Based Machine Learning. This work involves the use of mathematical programming methods to solve classification and other neural-net-style problems. Recent applications: credit scoring systems, bank failure prediction, and processing of spectral images for GIS and geoscience applications.
• Data Fusion. Merging data records from multiple sources to create statistically accurate composite data sets for research and modeling. Applications come from the U.S. Treasury Department, healthcare areas.
• Benchmarking with Data Envelopment Analysis We are developing DEA into a benchmarking methodology and non-statistical data mining technique. Research includes new DEA-based models, computationally efficient solution algorithms, processes for identifying performance factors and metrics for benchmarking studies, and industry-specific applications. This work is sponsored by the National Science Foundation. Applications include Pier 1 Imports and the Federal Reserve Banking System.
• Optimization Alternatives for Data Mining. Included are formulations and implementations of alternatives to the standard multivariate methodologies such as various forms of regression, principle components analysis, Lasso, and discriminant analysis
   
• Optimization for Quantum and High-Performance Computers. This work focuses on the development of optimization algorithms (both exact and inexact) that can exploit multiple processors and adiabatic quantum computers to reduce problem solution time. Applications include branch-and-bound fixed charge networks and large-scale DEA.
   
• Empirical Analysis of Algorithms. Designing and conscientiously reporting on computational experiments with algorithms has always been a challenge for computer scientists and, particularly, operations researchers. This research explores the issues and suggests means by which the CS/OR community can be more scientific in its approach to this valuable process.

Meet my academic ancestors Poisson, Lagrange, Euler, Bernoulli, and Laplace at the Mathematics Genealogy Project. My Erdös Number is 3. (Barr > Glover > Colbourn > Erdös).