Guang Cheng: A Profile of Research and Reviews at UCLA
Guang Cheng is a professor at UCLA whose research focuses on generative data science, deep learning theory, and statistical machine learning. He received his BA in Economics from Tsinghua University in 2002 and PhD in Statistics from the University of Wisconsin-Madison in 2006. His work extends to the Trustworthy AI Lab at UCLA, which envisions AI 2.0 as being driven by trustworthiness, surpassing mere performance, and being built upon generative data, enhancing raw data. The lab develops data-centric tools, such as artificially generated tables, which enable privacy-preserving data sharing, with applications in digital marketing, finance, and healthcare sectors.
Research and Academic Contributions
Cheng's research interests are primarily centered around the burgeoning field of generative data science. His work in the Trustworthy AI Lab involves developing data-centric tools, such as artificially generated tables, which enable privacy-preserving data sharing, with applications in digital marketing, finance, and healthcare sectors. Cheng's work also includes synthetic audio data, generative data theory, and tensor graphical models.
Mentorship and Collaboration
Professor Cheng has advised and co-advised numerous graduate students and postdoctoral researchers, contributing to a wide array of research areas:
- Peiyu Yu: Statistical-and-Computational Trade-off in Big Data.
- Shih-Kang Chao: Distributed and Online Statistical Inference.
- Jincheng Bai: Sparse Deep Neural Networks.
- Pratik Ramprasad: Policy Evaluation in Statistical Reinforcement Learning.
- Yue Xing: Statistical Theory for Adversarial Robustness in Machine Learning.
- Zhanyu Wang: Statistical Inferences with Differential Privacy Guarantee.
- Marie Maros: Decentralized High-dimensional Statistical Learning.
- Yuantong Li: Statistical Matching Model in Centralized Two-sided Markets.
- Wei Sun and Chendi Wang: Generative Data with Privacy Guarantees.
- Xianli Zeng: Statistical Foundation of Fairness.
- Jacob Swoveland: The Impact of Generative and Real Training Data on Model Vulnerability.
- Ryan O'Dell: Estimating Privacy Leakage of Machine Learning Models.
- Nicklaus Kim: Privacy Auditing of Generative Tabular Data Generators Using Membership Inference Attacks.
The Trustworthy AI Lab at UCLA is actively seeking Master, PhD, and Postdoctoral researchers who possess a strong passion for Trustworthy AI, Generative Data Science, and AI Agents. Undergraduate students majoring in statistics, computer science, or mathematics are welcome to apply for the summer internship program. Past summer interns have gone on to attend top graduate schools, such as Stanford, Berkeley and Princeton.
Student Reviews and Concerns
Student reviews paint a mixed picture of Professor Cheng's teaching abilities, with a recurring theme of difficulty in his course structuring and communication.
Read also: Fact vs. Misinformation: Kuen Cheng Incident
Course Structure and Content Delivery
Several students have expressed concerns about the structure of Professor Cheng's courses. Some reviews mention that the initial weeks of the course are taught at a slow pace, covering content that is relatively easy. However, the course difficulty increases significantly in the second half, with a rapid pace through more complex topics. One student noted that Professor Cheng rushed through variance, covariance, and MGFs in the last three lectures before the final exam. Additionally, he was occasionally absent, leaving a TA to teach important concepts.
Communication and Teaching Style
A significant number of reviews highlight communication issues. Some students mention a language barrier due to Professor Cheng's accent. Others claim that he does not answer student questions properly, often responding with "I don't knows", non-answers, and wrong answers. Students have also expressed frustration with his theoretical lectures, which they find unhelpful for understanding the homework assignments.
Exam and Grading Issues
Several reviews mention problems with exams and grading. One student recounted a final exam where the average score was around 50% and was not curved. Another described a chaotic final exam situation where the TA arrived late with an insufficient number of copies, and the exam had to be cut short due to another class needing the room. Some students felt that Professor Cheng showed a lack of care about the midterm and final, evidenced by his absence.
Availability and Support
Students have also reported difficulties in contacting Professor Cheng for help. Some reviews state that he is often unavailable during his office hours and that he refuses to set up appointments unless they are weeks in advance. Students feel that he is primarily focused on research and does not prioritize teaching or student support.
Positive Aspects
Despite the numerous criticisms, some students acknowledge that Professor Cheng explains things clearly during the first few weeks of the course. Others recognize his deep understanding of the subject matter.
Read also: UCLA vs. Illinois: Basketball History
Read also: Navigating Tech Breadth at UCLA
tags: #Guang #Cheng #UCLA #professor
