BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Drupal iCal API//EN X-WR-CALNAME:Events items teaser X-WR-TIMEZONE:America/Toronto BEGIN:VTIMEZONE TZID:America/Toronto X-LIC-LOCATION:America/Toronto BEGIN:DAYLIGHT TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:20220313T070000 END:DAYLIGHT BEGIN:STANDARD TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 DTSTART:20221106T060000 END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:682e21112ee43 DTSTART;TZID=America/Toronto:20221202T120000 SEQUENCE:0 TRANSP:TRANSPARENT DTEND;TZID=America/Toronto:20221202T120000 URL:/combinatorics-and-optimization/events/combinatoria l-optimization-reading-group-jacob-skitsko SUMMARY:Combinatorial Optimization Reading Group - Jacob Skitsko CLASS:PUBLIC DESCRIPTION:Summary \n\nTITLE: Boosted Sampling\n\nSpeaker:\n Jacob Skitsko \n\nAffiliation:\n University of À¶Ý®ÊÓÆµ\n\nLocation:\n MC 6029 or contac t Rian Neogi for Zoom link\n\nABSTRACT: We will discuss the boosted samp ling technique introduced by\nGupta et al. which approximates the stochast ic version of problems by\nusing nice approximation algorithms for the det erministic version of\nthe problem. We will focus on rooted stochastic Ste iner trees as an\nexample\, though other problems are covered by this appr oach (such as\nvertex cover and facility location). The problem is given t o us in two\nstages: in the first stage we may choose some elements at a c heaper\ncost\, and in the second stage our actual requirements are reveale d to\nus\, and we can buy remaining needed elements at a more expensive co st\n(where costs get scaled by some factor in the second stage). We will\n see that if our problem is sub-additive\, and we have an\nalpha-approximat ion algorithm for the deterministic version of our\nproblem with a beta-st rict cost-sharing function then we can get an\n(alpha + beta)-approximatio n for the stochastic version of our\nproblem. We also discuss related prob lems\, for example the (not\nsub-additive!) unrooted stochastic Steiner tr ee problem.\n DTSTAMP:20250521T185305Z END:VEVENT END:VCALENDAR