Stefan Wild is currently developing model-based methods for derivative-free optimization that exploit additional structures often found in practical applications, including constraints, discrete variables, computational noise, parallel computing environments, and parameter estimation. In addition to numerical optimization, he is interested in machine learning and numerical linear algebra.
Wild joined LANS as an Argonne Director's Postdoctoral Fellow in September 2008. He obtained his Ph.D. in operations research at Cornell University.
- Algorithms and software for simulation-based optimization problems where derivatives of the objective are unavailable