BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//University of Liverpool Computer Science Seminar System//v2//EN
BEGIN:VEVENT
DTSTAMP:20260922T052436Z
UID:Seminar-dept-467@lxserverM.csc.liv.ac.uk
ORGANIZER:CN=Lutz Oettershagen:MAILTO:Lutz.Oettershagen@liverpool.ac.uk
DTSTART:20181012T130000
DTEND:20181012T140000
SUMMARY:School Seminar Series
DESCRIPTION:Dr. Takanori Maehara: Stochastic Probing with Prices\n\nA recent trend in combinatorial optimization is Optimization with\n\nUncertainty. Here, we consider the following setting, called\n\nStochastic Probing with Prices.\n\nSuppose that we have a combinatorial optimization on a finite set V.\n\nEach element u in V is active (with probability p) or non-active\n\n(otherwise), which is determined by nature. By paying cost c_u, we can\n\nobserve whether u is active or not (called "probe"), and if it is\n\nactive, it is irrevocably added the solution. The goal is to maximize\n\nan objective function minus the payment, where the solution (= set of\n\nprobed active elements) and the set of probed elements satisfy their\n\nconstraints.\n\nThis problem is a generalization of the Stochastic Probing\n\n(Gupta-Nagarajan, IPCO'13) that does not contain the cost, and the\n\nPrice of Information (Singla, SODA'18) that does have constraint on\n\nprobed elements. This problem has several applications including\n\nkidney exchange, online dating, and online advertising.\n\nIn this study, we propose a method to obtain approximate strategy when\n\nthe objective function is submodular and the constraints have low\n\ncorrelation gaps. The method is based on the continuous greedy\n\nalgorithm and the contention resolution scheme.\n\nThis is a joint work with Ben Chugg (UBC)\n\n\n\nShort bio:\n\n\n\nDr. Takanori Maehara is a Unit Leader (= Associate Professor) of Discrete\n\nOptimization Unit at RIKEN AIP, Japan. He completed his PhD at the\n\nUniversity of Tokyo in 2012. He works on discrete optimization theory,\n\nnumerical analysis, and these applications in machine learning and\n\ndata mining.\n\nhttps://www.csc.liv.ac.uk/research/seminars/abstract.php?id=467
LOCATION:H223
END:VEVENT
END:VCALENDAR