BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Asia/Singapore BEGIN:DAYLIGHT DTSTART:20380119T111407 TZOFFSETFROM:+0800 TZOFFSETTO:+0800 RRULE:FREQ=YEARLY;BYDAY=3TU;BYMONTH=1 TZNAME:+08 END:DAYLIGHT BEGIN:STANDARD DTSTART:19820101T000000 TZOFFSETFROM:+0730 TZOFFSETTO:+0800 RRULE:FREQ=YEARLY;BYDAY=1FR;BYMONTH=1 TZNAME:+08 END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260324T030229Z UID:965110DD-6D72-4768-AB79-7143CB15B69A DTSTART;TZID=Asia/Singapore:20260317T150000 DTEND;TZID=Asia/Singapore:20260317T170000 DESCRIPTION:Evolutionary algorithms mimic natural processes such as selecti on\, crossover\, and mutation to solve a wide range of optimization proble ms. Effective application of these algorithms requires a solid understandi ng of the various selection\, crossover\, and mutation operators. However\ , exploration and exploitation are fundamental processes—and arguably th e most critical concepts—for any search algorithm. Despite their importa nce\, these concepts are not yet well understood among many researchers an d practitioners. Furthermore\, the direct measurement of exploration and e xploitation remains an open problem in evolutionary computation.\n\nThis t alk first introduces the basic components of evolutionary algorithms\, hig hlights common issues and mistakes encountered by inexperienced users\, an d discusses their diverse applications. In the second part\, a novel direc t measure of exploration and exploitation based on attraction basins is pr esented. Attraction basins are regions of a search space in which each reg ion contains a point\, called an attractor\, toward which neighbouring poi nts tend to evolve. Each search point can therefore be associated with a s pecific attraction basin. If a newly generated search point belongs to the same attraction basin as its parent\, the search process is considered ex ploitation\; otherwise\, it is classified as exploration.\n\nIn the final part\, the talk demonstrates how the newly developed exploration and explo itation measures can be used to analyze and compare different evolutionary algorithms.\n\nEngineering Auditorium\, National University of Singapore\ , Singapore\, Singapore\, Singapore LOCATION:Engineering Auditorium\, National University of Singapore\, Singap ore\, Singapore\, Singapore ORGANIZER:dipti@nus.edu.sg SEQUENCE:8 SUMMARY:Direct Measuring of Exploration and Exploitation in Evolutionary Al gorithms using Attraction Basin-based Measures URL;VALUE=URI:https://events.vtools.ieee.org/m/546416 X-ALT-DESC:Description: <br /><p>Evolutionary algorithms mimic natural proc esses such as selection\, crossover\, and mutation to solve a wide range o f optimization problems. Effective application of these algorithms require s a solid understanding of the various selection\, crossover\, and mutatio n operators. However\, exploration and exploitation are fundamental proces ses&mdash\;and arguably the most critical concepts&mdash\;for any search a lgorithm. Despite their importance\, these concepts are not yet well under stood among many researchers and practitioners. Furthermore\, the direct m easurement of exploration and exploitation remains an open problem in evol utionary computation.</p>\n<p><br>This talk first introduces the basic com ponents of evolutionary algorithms\, highlights common issues and mistakes encountered by inexperienced users\, and discusses their diverse applicat ions. In the second part\, a novel direct measure of exploration and explo itation based on attraction basins is presented. Attraction basins are reg ions of a search space in which each region contains a point\, called an a ttractor\, toward which neighbouring points tend to evolve. Each search po int can therefore be associated with a specific attraction basin. If a new ly generated search point belongs to the same attraction basin as its pare nt\, the search process is considered exploitation\; otherwise\, it is cla ssified as exploration.</p>\n<p><br>In the final part\, the talk demonstra tes how the newly developed exploration and exploitation measures can be u sed to analyze and compare different evolutionary algorithms.</p> END:VEVENT END:VCALENDAR