Huang Fang

Huang Fang is a Research Scientist at Baidu Research's Cognitive Computing Lab, specializing in optimization theory, machine learning, and data mining. With a strong academic background from top institutions, Huang has established themselves as a prolific researcher in the field of machine learning, particularly in federated learning, privacy-preserving algorithms, and optimization methods.

• Ph.D. in Computer Science, 2021, University of British Columbia
• M.S. in Statistics and Computer Science, 2017, University of California, Davis
• B.Sc. in Financial Mathematics, 2015, Central University of Finance and Economics (China)

Huang Fang's research spans multiple critical areas in machine learning, with a particular focus on optimization theory and its applications. Their work bridges theoretical foundations with practical implementations, especially in federated learning systems where privacy and efficiency are paramount. The research portfolio demonstrates expertise in developing novel algorithms for large-scale learning problems, with particular emphasis on convergence analysis, sparse optimization, and scalable methods for knowledge representation. Huang has made significant contributions to understanding the theoretical properties of optimization algorithms under various constraints and settings.

The publication record reveals a strong trajectory in advancing machine learning methodologies, with recent work focusing on differentiable neuro-symbolic reasoning, efficient Markov logic networks, and improved convergence guarantees for privacy-preserving stochastic optimization. The research shows a consistent pattern of addressing fundamental challenges in distributed and large-scale machine learning, particularly in the areas of federated learning fairness, efficient coordinate descent methods, and matrix completion techniques.

Huang Fang has demonstrated exceptional productivity in high-impact venues including NeurIPS, ICML, ICLR, WWW, and other top-tier machine learning conferences. The work consistently combines theoretical rigor with practical applicability, making significant contributions to both the foundational understanding of optimization methods and their implementation in real-world systems.