Hi there, I'm Yue ZHAO (赵越 in Chinese)! 👋

I am a Ph.D. student at Carnegie Mellon University (CMU), a machine learning (ML) evangelist, and an ex management consultant at PwC Canada. As a seasoned ML software/system architect, I have led/participated > 10 ML libraries initiatives, 8,000 GitHub stars (top 0.002%: ranked 820 out of 40M GitHub users), and >200,0000 total downloads.

Good news: I am looking for 2021 Summer ML/DM Internship in Canada, United States, or China. Not necessarily pure research; system or AutoML related stuff would be great fit as well. Just reach out and we could work something out :)

And of course, I am still a ML/DM researcher at the end of the day.

🔭 Reseach-wise, I mainly focus on:

• data mining topics related to scalability, reliability, and automation and
• information systems questions related to interaction, trade-off, and cooperation between human and “AI”

😄 I am open to:

• collaboration opportunities (anytime & anywhere & any type) and
• research internships (open for Summer 2021). I could legally work in Canada, United States, and China

📫 Contact me by:

• Email (zhaoy [AT] cmu.edu)
• Twitter
• 知乎:「微调」
• WeChat (微信)

💬 News & Travel:

• Oct 2020: Our demo paper TODS: An Automated Time Series Outlier Detection System is accepted at AAAI 2021. See the new Python library TODS (video) designed for time-series outlier detection. It is initialized and led by DATA Lab @ Texas A&M University; I contribute to core detection model design and implementation.

• Oct 2020: Our paper AutoAudit: Mining Accounting and Time-Evolving Graphs will appear in BigData 2020. It is a system for detecting anomalies in time-evolving graphs. Great work with Meng-Chieh Lee, Aluna Wang, Pierre Jinghong Liang, Leman Akoglu, Vincent S. Tseng, and Christos Faloutsos!

• Oct 2020: Have a new system paper (SUOD: Accelerating Large-scare Unsupervised Heterogeneous Outlier Detection) under review at a major ML conference. SUOD is an acceleration system for large-scale unsupervised outlier detection. It has been downloaded by more than 700,000 times, included as part of PyOD, and presented in AAAI Workshop on Artificial Intelligence for Cyber Security (AICS).

• Sep 2020: We have a new paper Automating Outlier Detection via Meta-Learning (code) out. In this paper, we propose the first unsupervised meta-learner that can select (recommend) the most performing outlier detection model on an arbitrary dataset.

• Sep 2020: Our paper COPOD: Copula-Based Outlier Detection (camera-ready version) will appear in ICDM 2020 soon! It is a fast, parameter-free, and highly interpretable unsupervised outlier detection algorithm and available in PyOD now.

• Sep 2020: Have a paper accepted at ICDM Workshops 2020 (ICDMW). A personal copy can be found here: SynC: A Copula based Framework for Generating Synthetic Data from Aggregated Sources.