Yilun Jin

Yilun Jin

PhD student at CSE, HKUST. Machine Learning/Artificial Intelligence

Welcome to Yilun Jin’s homepage!

I am currently a Ph.D. student at the CSE Department, the HKUST, advised by Prof. Qiang Yang and co-advised by Prof. Kai Chen. My research interests lie in data mining algorithms for practical problems, including:

  • Expressive Learning Models
    • Structure-Aware Models for Graph Data [GraLSP, AAAI’20; GraphSTONE, KDD’20; GraphMSE, AAAI’21; GSKN, WWW’21]
    • Temporal-Aware Models under Dynamic Data [EPNE, ECAI’20]
  • Effective Learning under the Lack of Data
    • Transfer Learning [TransGTR, KDD’23; CrossTReS, KDD’22; MUSDAC, IJCAI’20]
    • Federated Learning [SecureBoost, IEEE Intel. Sys. 2021; GeniBatch, EuroSys’24]

Prior to joining the HKUST, I spent 4 wonderful years at Peking University, China, obtaining both a B.S in Computer Science and a B.Economics at the National School of Development. I was fortunate to have Prof. Guojie Song as my advisor during my days at PKU, with whom I collaborate very well.

You can contact me via (yilun.jin [at] connect.ust.hk).

I am on the job market for 2024 and 2025, looking for industrial positions on large language models (LLM), domain-specific LLMs, transfer learning, and graph machine learning. Drop me a message if you are interested!

News

[September 2024] Shopping MMLU, the dataset used to organize Amazon KDD Cup 2024, is accepted at NeurIPS 2024 Datasets and Benchmarks Track. Congratulations!

[March 2024] We are hosting the 🛍️Amazon KDD Cup 2024: Multi-Task Online Shopping Challenge for LLMs with plenty of awards. Click here to contribute ingenious solutions 🚀!

[September 2023] Beginning my internship at Amazon Rufus Team!

[August 2023] One co-authored paper is accepted at EuroSys 2024. Congratulations!

[May 2023] One paper is accepted at KDD 2023. Congratulations!

[February 2023] One co-authored paper is accepted at SIGMOD 2023. Congratulations!

[December 2022] Invited to serve on the Program Committee for KDD 2023 for the first time!

[December 2022] SecureBoost received the 2021 Best Paper Award for IEEE Intelligent Systems. Congratulations!

[October 2022] Recognized as a Top Reviewer of NeurIPS 2022!

[July 2022] Recognized as an Outstanding Reviewer (Top 10%) of ICML 2022!

[May 2022] CrossTReS is accepted at KDD 2022. Congratulations!

[August 2021] I passed my Ph.D. qualifying exam and became a Ph.D. candidate. [slides]

[May 2021] SecureBoost is accepted at IEEE Intelligent Systems. Congratulations to my co-authors!

[January 2021] One co-authored paper is accepted at TheWebConf 2021 as a long paper. Congratulations to my co-authors!

[December 2020] One co-authored paper is accepted at AAAI 2021. Congratulations to my co-authors!

[May 2020] One co-authored paper accepted at KDD 2020. Congratulations to my co-authors!

[January 2020] Two papers accepted at TheWebConf 2020 (Short Paper) and ECAI 2020.

[November 2019] GraLSP is accepted at AAAI 2020 [Paper]. Congratulations!

Miscellaneous

I hold a broad range of interests so as to keep enjoying life.

I enjoy sports including jogging and basketball. My favourite team in NBA is San Antonio Spurs and my favourite football club is Liverpool FC.

I am a civil aviation enthusiast. I am a member of the ACICFG, a team devoted to making subtitles and captions for documentaries related to civil aviation (such as the Air Crash Investigation). Some of our works are published on bilibili. I also write some articles on zhihu related to civil aviation and my research.

I am an alcoholist, I like the following types of alcohol.

  • Beers, including Pale Ale (I like Ballast Point, Brooklyn, Dogfish and BrewDog.), Trappist (I like Rochefort and Chimay.) and Hefeweizen (I think Franziskaner is good enough.).

  • Vodka. Smirnoff Black, Grey Goose and Russian Standards are of high quality.

  • Whiskey. For single malts, I like Talisker and Lagavulin. For blended, Johnnie Walker Double Black is worth a try.

  • Bourbon. I have only tried Maker’s Mark and I found it charming. Recently I tried Wild Turkey 101. I found it more straight and less mild than Maker’s Mark, but still interesting.

  • Gin. The taste of Gin is a highly personalized matter. I would recommend the Botanists and Monkey 47.

An image demonstrating genres of single malt scotch is here.

Friends

I make friends with many elite guys who are doing state-of-the-art research in many areas. Some of their homepages are below.

Haodong Duan (Researcher at Shanghai AI Laboratory, Computer Vision).

Peiqing Chen (Ph.D student at University of Maryland, Network measurement).