A Modified Sample Approximation Method for Chance Constrained Problems


Department of Systems Engineering and Engineering Management
The Chinese University of Hong Kong

Title: A Modified Sample Approximation Method for Chance Constrained Problems
Speaker: Dr. Jianqiang Cheng,
LRI, University of Paris 11
Date: July 29, 2014 (Tuesday)
Time: 4:30 PM to 5:30 PM
Venue: Room 513
       William M.W. Mong Engineering Building (ERB)
       (Engineering Building Complex Phase 2)
       The Chinese University of Hong Kong.

Short Biography of the speaker:
Jianqiang Cheng is a postdoctoral fellow in Department of Computer Science (LRI) at University of Paris 11, working on “Stochastic Nuclear Outage problems with joint chance constraints”. He completed his Ph.D. in 2013 at the same university with topic on stochastic combinatorial optimization. He received his Master degree and B.S. Degree in Math and Applied Maths in Shanghai University, with his master research in the area of nonparametric statistics. He is particularly interested in chance-constrained programming, semi-definite programming, as well as nonparametric statistics. Dr. Cheng has published the major part of his PhD and postdoc results in international leading journals and conferences.
Since chance constrained programming (CPP) was firstly introduced by Charnes, Cooper and Symonds in 1958, it has attracted significant attention of many researchers and practitioners as it plays an important role in engineering, telecommunication, finance, etc. However, little progress was made until recently because of two reasons. One main reason is that the feasible set of CPP is generally non-convex even if the inside function is affine function. The other reason is that checking feasibility of given solution is generally hard, as it requires multi-dimensional integrations. As chance- constrained problems are generally intractable, it leads to the development of solving methods in two directions. One is to apply convex (or tractable) approximations. The other approach is to use sampling methods to approximate original problems, such as sample average approximation (SAA).
In this talk, we focus on the sampling method to solve CPP. First of all, a modified sample average approximation is presented (MSAA). With the modified sample method, despite that some other chance constraints arise, the corresponding problem has no binary variable whereas there are binary variables based on the traditional sample average approximation. Second, we show that, for the new chance constraints, it is easy to handle these chance constraints in some cases. Third, we show the modified sample method has same convergence properties as SAA. Finally, numerical experiments are conducted to compare the proposed approximation to SAA in order to show the strength of the new sample method.
                      Everyone is invited to attend the talk.
The talk will be hosted by:
       Prof. Janny Leung,
       Department of Systems Engineering and Engineering Management,
       The Chinese University of Hong Kong,
       Telephone Number: (852) 3943-8238
For general enquiries, please contact the student coordinator:
       Andy Chung,
       Department of Systems Engineering and Engineering Management,
       The Chinese University of Hong Kong,
SEEM-5202 Website:
Tuesday, July 29, 2014 - 08:30 to 09:30