Geon-Hyeong Kim (김건형)

Email: ghkim@lgresearch.ai

Research Interest: Imitation Learning, Reinforcement Learning, Machine Learning

Experience

2022. 07. - Current. AI Scientist at LG AI Research

Education

2015. 03. - 2022. 08. PhD, School of Computing, KAIST, Korea (Advisor: Kee-Eung Kim, Hongseok Yang)

2013. 03. - 2015. 02. MS, Mathematical Sciences, KAIST, Korea (Advisor: Chang-Ock Lee)

2009. 02. - 2013. 02. BS, Mathematical Sciences, KAIST, Korea (Minor in School of Computing)

Publications

International

[C8] LobsDICE: Offline Learning from Observation via Stationary Distribution Correction Estimation

  • Geon-Hyeong Kim*, Jongmin Lee*, Youngsoo Jang, Hongseok Yang, Kee-Eung Kim (*: equal contribution)

  • NeurIPS 2022 (to appear)

[C7] DemoDICE: Offline Imitation Learning with Supplementary Imperfect Demonstrations

  • Geon-Hyeong Kim, Seokin Seo, Jongmin Lee, Wonseok Jeon, HyeongJoo Hwang, Hongseok Yang, Kee-Eung Kim

  • ICLR 2022

[C6] Multi-View Representation Learning via Total Correlation Objective

  • HyeongJoo Hwang, Geon-Hyeong Kim, Seunghoon Hong, Kee-Eung Kim

  • NeurIPS 2021

[C5] Variational Interaction Information Maximization for Cross-domain Disentanglement

  • HyeongJoo Hwang, Geon-Hyeong Kim, Seunghoon Hong, Kee-Eung Kim

  • NeurIPS 2020

[C4] Variational Inference for Sequential Data with Future Likelihood Estimates

  • Geon-Hyeong Kim, Youngsoo Jang, Hongseok Yang, Kee-Eung Kim

  • ICML 2020

[C3] Monte-Carlo Tree Search in Continuous Action Spaces with Value Gradients

  • Jongmin Lee, Wonseok Jeon, Geon-Hyeong Kim, Kee-Eung Kim

  • AAAI 2020

[C2] Trust Region Sequential Variational Inference

  • Geon-Hyeong Kim, Youngsoo Jang, Jongmin Lee, Wonseok Jeon, Hongseok Yang, Kee-Eung Kim

  • ACML 2019

[C1] Monte-Carlo Tree Search for Constrained POMDPs

  • Jongmin Lee, Geon-Hyeong Kim, Pascal Poupart, Kee-Eung Kim

  • NeurIPS 2018

[W1] Monte-Carlo Tree Search for Constrained MDPs

  • Jongmin Lee, Geon-Hyeong Kim, Pascal Poupart, Kee-Eung Kim,

  • ICML Workshop on Planning and Learning (PAL-18), 2018

[J1] Bayesian optimistic Kullback–Leibler exploration

  • Kanghoon Lee, Geon-Hyeong Kim, Pedro Ortega, Daniel D. Lee, Kee-Eung Kim

  • Machine Learning, Springer, 2018

Academic Talks

DemoDICE: Offline Imitation Learning with Supplementary Imperfect Demonstrations

  • 2022. 04. 25 - 29, ICLR (virtual)

Variational Inference for Sequential Data with Future Likelihood Estimates

  • 2020. 07. 16 - 17, ICML (virtual)

  • 2019. 09. 27, Kakao, Korea

Trust Region Sequential Variational Inference

  • 2019. 11. 19, ACML

Reviewer

  • ACML (2019, 2020)

  • ICLR (2021, 2022)

  • ICML (2021, 2022)

  • IJCAI (2021, 2022)

  • NeurIPS (2020, 2021)

Teaching Experience

  • KAIST-Samsung AI Expert Program: Introduction to Reinforcement Learning, TA at KAIST (2020)

  • Data Structures (CS206), TA at KAIST (2016 Fall, 2018 Spring)

  • Introduction to Artificial Intelligence (CS470), TA at KAIST (2017 Fall)

  • Introduction to Linear Algebra (MAS109), TA at KAIST (2013 Fall, 2014 Spring, 2014 Fall)