Kwang-Sung Jun
Scholar

Kwang-Sung Jun

Google Scholar ID: VgvC7o8AAAAJ
The University of Arizona
Machine LearningMulti-armed banditonline learningactive learning
Citations & Impact
All-time
Citations
1,565
 
H-index
18
 
i10-index
29
 
Publications
20
 
Co-authors
32
list available
Resume (English only)
Academic Achievements
  • Jan'25: 2 papers accepted to AISTATS;
  • Nov'24: Gave a talk at UW-Madison SILO on 'Confidence Sequences via Online Learning';
  • Oct'24: 2 papers accepted to NeurIPS;
  • May'24: 2 papers accepted to ICML;
  • May'24: 1 paper accepted to COLT;
  • Feb'24: Will serve as an action editor for the journal Machine Learning;
  • Jan'24: 1 paper accepted to AISTATS;
  • Dec'23: 1 paper accepted to NeurIPS.
Research Experience
  • Before joining the University of Arizona, he was a postdoctoral researcher at Boston University; obtained his PhD from the University of Wisconsin-Madison and held a postdoctoral position at the Wisconsin Institute for Discovery.
Education
  • PhD: University of Wisconsin-Madison, Advisor: Xiaojin (Jerry) Zhu;
  • Postdoc: Boston University, Advisor: Francesco Orabona;
  • Postdoc: Wisconsin Institute for Discovery, Collaborators: Robert Nowak, Rebecca Willett, Stephen Wright.
Background
  • Research interests: Interactive Machine Learning (IML), including reinforcement learning, bandits, Bayesian optimization, and active learning. Also develops novel confidence bounds that often become key tools for constructing efficient IML algorithms. Recently, he has been looking into IML problems arising from GenAIs, including LLMs.