Seungwon Kim

I am an electrical engineer at Incheon International Airport Corporation. I have been working at Incheon Airport Corporation as an electrical engineer since 2015. I have sought to advance the airport’s electrical signal data collection by managing the SCADA system. After I recognized the potential impact AI can have on electrical engineering and the airport overall, I enrolled in Georgia Tech's OMSCS program and successfully completed my master's in computer science in 2020 while working full time.

My research interests lie in the field of Reinforcement Learning and Natural Language Processing. I have worked on Batch Reinforcement Learning, Text Summarization, and Adapters, resulting in workshop publications at EMNLP and ACL.

I'm deeply passionate about machine learning and want to utilize AI to improve my work as well as the quality of life for all and positively impact society.

Email  /  CV  /  Github  /  LinkedIn

Revisiting Pretraining with Adapters
Seungwon Kim, Alex Shum, Nathan Susanj, Jonathan Hilgart
ACL Workshop on Representation Learning for NLP, 2021, Best Paper Award

Using Pre-Trained Transformer for Better Lay Summarization
Seungwon Kim
EMNLP Scholarly Document Processing Workshop, 2020

• Neurips 2019 Reproducibility Challenge: I participated in Neurips Reproducibility Challenge, reproduced experiments in Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction , and wrote reproducibility report. [Code] [Report]
• Striving for Simplicity in Off-policy Deep Reinforcement Learning: I implemented variants of Deep Q-Networks in Striving for Simplicity in Off-policy Deep Reinforcement Learning including classic DQN , C51 , QR-DQN , Ensemble DQN and REM DQN. [Code]

Template from Jon Barron's website.