Reinforcing AI and Network
RAIN Lab에서는 인공지능(AI), 기계 학습(ML), 그리고 네트워크를 중심으로 첨단 시스템 연구를 수행하고 있습니다. 현대 시스템은 정보의 파편화와 분산화가 특징이며, 복잡한 네트워크 구조 속에서 필요한 정보를 신속하고 정확하게 수집하고 분석하는 능력이 필수적입니다. 본 연구실은 이러한 분산 시스템의 효율적 활용을 목표로 Networked AI/ML 기술을 연구하고 있으며, AI/ML 파이프라인 자동화(AutoML)와 모델 경량화 등 다양한 연구를 통해 AI/ML 기술의 효율성과 성능 향상에 기여하고 있습니다.
The RAIN Lab conducts cutting-edge research focused on artificial intelligence (AI), machine learning (ML), and networks. Modern systems are characterized by fragmented and decentralized information, making it essential to collect and analyze the necessary data swiftly and accurately within complex network structures. Our lab focuses on developing Networked AI/ML technologies to efficiently utilize such decentralized systems, while also exploring various approaches to enhance the efficiency and performance of AI/ML, including the automation of AI/ML pipelines (AutoML) and model optimization.
RAIN Lab에서는 AI/ML 및 네트워크에 대한 연구에 관심이 있는 학부생 인턴과 석/박사 대학원생을 모집하고 있습니다. 특히 아래 주제들에 대해 관심이 있으시면 주창희 교수님께 연락을 주시기 바랍니다. (이메일: changhee@korea.ac.kr , 연구실: 고려대학교 우정정보관 507B)
Research on networks for AI services, jointly optimizing AI computing and networking
Research on AI-based automation - Automated Machine Learning (AutoML), Neural Architecture Search (NAS)
Recent news:
[2024.09] Byoungkyu Ji, Daeun Kim, Harim Kang are joining our lab. Welcome aboard!
[2024.04] Our proposal for 5-year project "Development of 6G APIs guaranteeing application performance through interaction with cellular systems" has been accepted. We are teamed with excellent researchers from SNU, UNIST, Hanyang Univ., and SKKU!
[2024.04] Our paper has been accepted at CVPR-NAS Workshop 2024.
Taegun An and Changhee Joo*, "CycleGANAS: Differentiable Neural Architecture Search for CycleGAN," CVPR-NAS Workshop, June 2024.
[2024.02] Jihyeon received her Ph.D degree after her academic journey. Many congratulations! She will remain in our lab for a short time and then move forward for her career!
[2024.02] Kukjin received his MS degree, and joined LG Electronics. Many congratulations!
[2024.02] Jihyeon Yun won the Best Presentation Award from A3 Foresight Program 2024 Workshop
[2024.01] Prof. Joo will serve as a General Co-Chair of WiOpt 2024, which will be held in Korea Univ.
[2023.12] Prof. Joo has been invited as a TPC member of ACM MobiHoc 2024.
[2023.12] Our paper has been accepted for publication in Elsevier Neurocomputing (SCI(E), IF 5.779).
Haesung Jo and Changhee Joo*, "AutoGAN-DSP: Stabilizing GAN Architecture Search with Deterministic Score Predictors," Elsevier Neurocomputing, accepted for publication.
[2023.11] Our paper has been accepted for publication in IEEE/ACM Trans. on Networking (ToN) (SCI(E), IF 3.7), which is the top journal in the network research area.
Sunjung Kang, Atilla Eryilmaz, and Changhee Joo*, "Comparison of Decentralized and Centralized Update Paradigms for Distributed Remote Estimation," IEEE/ACM Transactions on Networking (ToN), accepted for publication.
[2023.11] Our paper has been accepted at NeurIPS 2023 Workshop Foundation Models for Decision Making (FMDM).
Kuk Jin Kim and Changhee Joo*, "Agnostic Architecture for Heterogeneous Multi-Environment Reinforcement Learning," NeurIPS 2023 Workshop FMDM (Foundation Models for Decision Making).