I am Quan Zhang, a Ph.D. candidate in Department of Information, Risk and Operation Management, McCombs School of Business, University of Texas at Austin.
Appreciating both statistics and model machine learning techniques, I am committed to bridging the gap between
the two seemingly separated areas.
My research is focused on interpretable learning to achieve a good balance of interpretability and model capacity.
I have invented a general Weibull racing framework that can be applied to interpretable nonlinear (hierarchical) classification,
survival analysis, learning to rank and regression for skewed and heterogeneous data.
I have applied it to the analysis of online borrowers' loan payoff and default, and demonstrated the value of
non-standard information that is often overlooked by traditional financial institutes.
Currently, I am working on variational inference for uncertainty estimation in big data applications.
I am a comedian and perform stand-up comedy which is called crosstalk, or Xiangsheng. If only one thing can be written in my resume, it must be that I performed Xiangsheng for 3,000 audiences at Peking University 2012 graduation party.
Methodology: Statistics, machine learning, interpretable learning, Bayesian methods, variational inference, big data.Modeling: Survival analysis, classification, discrete choice models, regression (for skewed and heterogeneous data). Application: Online finance (particularly crowdfunding), missing data imputation, bioinformatics, medical data analysis, clinical trial
Quan Zhang and Mingyuan Zhou. "Improving Variational Inference by Adaversarial Learning and Reparameterizable Updating Equations."Quan Zhang, Qiang Gao, Mingfeng Lin and Mingyuan Zhou. "Weibull Racing Time-to-event Modeling and Analysis of Online Borrowers' Loan Payoff and Default." Submitted to 2020 GSU-RFS FinTech Conference.
Quan Zhang and Mingyuan Zhou. "Nonparametric Bayesian Lomax Racing for Survival Analysis with Competing Risks." Advances in Neural Information Processing Systems. 2018. codeQuan Zhang and Mingyuan Zhou. "Permuted and Augmented Stick-Breaking Bayesian Multinomial Regression." Journal of Machine Learning Research (2018): Vol. 18(204) 1−33. Quan Zhang, Youssef Toubouti, and Bradley P. Carlin. "Design and analysis of Bayesian adaptive crossover trials for evaluating contact lens safety and efficacy." Statistical Methods in Medical Research (2017): Vol. 26(3) 1216–1236.