- Seminar Calendar
- Seminar Archive
- 2024-2025 Semester 1
- 2023-2024 Semester 2
- 2023-2024 Semester 1
- 2022-2023 Semester 2
- 2022-2023 Semester 1
- 2021-2022 Semester 2
- 2021-2022 Semester 1
- 2020-2021 Semester 2
- 2020-2021 Semester 1
- 2019-2020 Semester 2
- 2019-2020 Semester 1
- 2018-2019 Semester 2
- 2018-2019 Semester 1
- 2017-2018 Semester 2
- 2017-2018 Semester 1
- 2016-2017 Semester 2
- 2016-2017 Semester 1
- 2015-2016 Semester 1
- 2015-2016 Semester 2
- 2014-2015 Semester 2
- 2014-2015 Semester 1
- 2013-2014 Semester 2
- 2013-2014 Semester 1
- 2012-2013 Semester 2
- 2012-2013 Semester 1
- 2011-2012 Semester 2
- 2011-2012 Semester 1
- 2010-2011 Semester 2
- 2010-2011 Semester 1
- 2009-2010 Semester 2
- 2009-2010 Semester 1
- 2008-2009 Semester 2
- 2008-2009 Semester 1
- 2007-2008 Semester 2
- 2007-2008 Semester 1
- 2006-2007 Semester 2
- 2006-2007 Semester 1
- 2005-2006 Semester 2
- 2005-2006 Semester 1
- Contact
- Site Map
A General Framework for Optimal Data-Driven Optimization
----------------------------------------------------------------------------------------------------
Faculty Distinguished Lecture
Faculty of Engineering
The Chinese University of Hong Kong
----------------------------------------------------------------------------------------------------
Title : A General Framework for Optimal Data-Driven Optimization
Speaker : Professor Daniel Kuhn, Professor of Operations Research, College of Management of Technology, EPFL
Date : Friday, 2 February 2024
Time : 2:00p.m. - 3:00p.m.
Venue : TY Wong Hall, 5/F, Ho Sin Hang Engineering Building
Abstract
We propose a statistically optimal approach to construct data-driven decisions for stochastic optimization problems. Fundamentally, a data-driven decision is simply a function that maps the available training data to a feasible action. It can always be expressed as the minimizer of a surrogate optimization model constructed from the data. The quality of a data-driven decision is measured by its out-of-sample risk. An additional quality measure is its out-of-sample disappointment, which we define as the probability that the out-of-sample risk exceeds the optimal value of the surrogate optimization model. The crux of data-driven optimization is that the data-generating probability measure is unknown. An ideal data-driven decision should therefore minimize the out-of-sample risk simultaneously with respect to every conceivable probability measure (and thus in particular with respect to the unknown true measure). Unfortunately, such ideal data-driven decisions are generally unavailable. This prompts us to seek data-driven decisions that minimize the out-of-sample risk subject to an upper bound on the out-of-sample disappointment - again simultaneously with respect to every conceivable probability measure. We prove that such Pareto-dominant data-driven decisions exist under conditions that allow for interesting applications: the unknown data-generating probability measure must belong to a parametric ambiguity set, and the corresponding parameters must admit a sufficient statistic that satisfies a large deviation principle. If these conditions hold, we can further prove that the surrogate optimization model generating the optimal data-driven decision must be a distributionally robust optimization problem constructed from the sufficient statistic and the rate function of its large deviation principle. This shows that the optimal method for mapping data to decisions is, in a rigorous statistical sense, to solve a distributionally robust optimization model. Maybe surprisingly, this result holds irrespective of whether the original stochastic optimization problem is convex or not and holds even when the training data is non-i.i.d. As a byproduct, our analysis reveals how the structural properties of the data-generating stochastic process impact the shape of the ambiguity set underlying the optimal distributionally robust optimization model.
Biography
Daniel Kuhn is Professor of Operations Research at the College of Management of Technology at EPFL, where he holds the Chair of Risk Analytics and Optimization (RAO). His current research interests are focused on data-driven optimization, the development of efficient computational methods for the solution of stochastic and robust optimization problems and the design of approximation schemes that ensure their computational tractability. Before joining EPFL, Daniel Kuhn was a faculty member in the Department of Computing at Imperial College London (2007-2013) and a postdoctoral research associate in the Department of Management Science and Engineering at Stanford University (2005-2006). He holds a PhD degree in Economics from University of St. Gallen and an MSc degree in Theoretical Physics from ETH Zurich.
Date:
Friday, February 2, 2024 - 14:00