BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Asia/Shanghai
BEGIN:STANDARD
DTSTART:19910915T010000
TZOFFSETFROM:+0900
TZOFFSETTO:+0800
TZNAME:CST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251110T040355Z
UID:B18E7027-ACE8-4682-982B-05515063356E
DTSTART;TZID=Asia/Shanghai:20251108T160000
DTEND;TZID=Asia/Shanghai:20251108T163000
DESCRIPTION:As modern optimization tasks grow in number and diversity\, aut
 omated algorithm design becomes essential. This talk introduces Meta-Black
 -Box Optimization (MetaBBO)\, from early ideas to recent advances\, and or
 ganizes the field into four paradigms: algorithm selection\, algorithm con
 figuration\, algorithm generation\, and solution manipulation. I will focu
 s on our contributions to algorithm generation along three directions: dis
 covering evolutionary update rules in a symbolic mathematical space\; asse
 mbling optimizers with parameters in a modular space\; and leveraging larg
 e language models to synthesize executable optimizer code. To strengthen g
 eneralization\, we study how to build diverse training task sets and how t
 o learn more effective task and state representations. I will also present
  the MetaBox platform\, which aims to improve the development efficiency a
 nd support fair evaluation in the field. I will end with a brief look ahea
 d to key opportunities and open questions.\n\nNo. 3333 Liuxian Avenue\, Na
 nshan District\, Shenzhen\, Guangdong\, China
LOCATION:No. 3333 Liuxian Avenue\, Nanshan District\, Shenzhen\, Guangdong\
 , China
ORGANIZER:ranchengcn@gmail.com
SEQUENCE:6
SUMMARY:The Past\, Present\, and Future Challenges of Meta-Black-Box Optimi
 zation
URL;VALUE=URI:https://events.vtools.ieee.org/m/510367
X-ALT-DESC:Description: <br /><p class="MsoNormal"><span lang="EN-US" style
 ="font-family: 'Times New Roman'\,serif\; mso-fareast-font-family: 宋体\
 ;">As modern optimization tasks grow in number and diversity\, automated a
 lgorithm design becomes essential. This talk introduces Meta-Black-Box Opt
 imization (MetaBBO)\, from early ideas to recent advances\, and organizes 
 the field into four paradigms: algorithm selection\, algorithm configurati
 on\, algorithm generation\, and solution manipulation. I will focus on our
  contributions to algorithm generation along three directions: discovering
  evolutionary update rules in a symbolic mathematical space\; assembling o
 ptimizers with parameters in a modular space\; and leveraging large langua
 ge models to synthesize executable optimizer code. To strengthen generaliz
 ation\, we study how to build diverse training task sets and how to learn 
 more effective task and state representations. I will also present the Met
 aBox platform\, which aims to improve the development efficiency and suppo
 rt fair evaluation in the field. I will end with a brief look ahead to key
  opportunities and open questions. </span></p>
END:VEVENT
END:VCALENDAR