Xiao (Victor) Pu

PhD candidate in Statistics, Department of Mathematics, UCSD.

Industrial Experiences

In summer 2015, I was interning at Teradata Lab, where I implemented the famous machine learning algorithm, AdaBoost, on Teradata Aster's SQL-MapReduce system with Java.

In summer 2016, I was interning at Yahoo! Advertising and working with their data science team. Built a new CTR prediction model using Vowpal-Wabbit with distributed training on large advertising datasets. Improved AUC by more than 15% and increased the CTR by more than 20%.

Back to top

Research Interests

My master's research at University of Minnesota was about developing compression algorithms for quantum field theoretic calculations.

Currently´╝îI'm working on various topics in statistical machine learning. I'm particularly excited about sparse principle component analysis(SPCA), and feature selection in high-dimensional regression, classification and clustering.

See here for my publications.

Back to top


Prior to starting my PhD study in Statistics at UCSD Math Department, I obtained a M.S. degree (2013) in Physics at University of Minnesota with Professor John R. Hiller and a Bachelor's (2011) degree in Engineering at Shanghai Jiao Tong University.

Back to top


Office: APM 6446

Back to top