FPGA-based coprocessor for text string extraction
N.K. Ratha, A.K. Jain, et al.
Workshop CAMP 2000
Decision makers who use optimization technology to generate plans and schedules need to deal not only with high data volume, velocity, and variety, but also data veracity. This presents a significant challenge due to the uncertainty inherent in most data arising from, for example, approximations and aggregations, error in instrumentation, and predictions of volatile supply and demand patterns. Creating user-friendly optimization tools to manage this uncertainty involves many challenges: large quantities of data and scenarios, incomplete data, complex mathematical models for stochastic and robust optimization, and lack of user adoption, to name a few. In this paper, we describe how we address these challenges through research toward an Uncertainty Toolkit for decision optimization. This toolkit solicits information on the uncertain data, automatically generates models that incorporate the uncertainty, and includes visual analytics for comparing outcomes. We demonstrate quantitative benefits, in terms of financial performance and stability, and describe qualitative benefits, in terms of user adoption, for two commerce case studies.
N.K. Ratha, A.K. Jain, et al.
Workshop CAMP 2000
Michael C. McCord, Violetta Cavalli-Sforza
ACL 2007
Michael Ray, Yves C. Martin
Proceedings of SPIE - The International Society for Optical Engineering
Oliver Bodemer
IBM J. Res. Dev