Stewart Hu
*** ********* *** • Needham, MA *****
813-***-**** • *********@*****.***
Summary
A computational Econometrician working in the field of quantitative marketing, trying to bring to-
gether the two worlds of marketing science and high performance computation. Having over 10 years’
experiences in building back-end enterprise applications in C++, and stumbled onto quantitative mar-
keting while working on my Ph.D. in Bayesian Econometrics. With extensive experiences and training
in computer science, economics, and statistics, I am well-positioned to contribute in the exciting areas
of statistical/machine learning, predicative analytics, and econometric modeling.
Experience
• In4mation Insights Needham, MA
Senior Director, Marketing Science January 2011 – Present
– Operationalizing the cutting-edge models from the marketing and Economics literature by
utilizing the best technologies in High-Performance Computation.
– Building enterprise Bayesian software to analyze large scale scanner and survey data, includ-
ing Hierarchical linear, count and discrete choice models, Hierarchical Vector Autoregressive
(VAR) model, Hierarchical Hidden Markov Model (HMM), Customer Lifetime Value (CLV)
model, etc. (C++, MCMC).
– Parallelized Bayesian models for multi-core Linux boxes and clusters (Open MP, MPI)
– Built a revenue optimizer based on the open source NLP package Ipopt (C++, R).
– Built a calibration engine to calibrate the State Preferences estimated from survey studies
with the Revealed Preferences obtained from scanner data.
– Ran a Marketing Mix Model to analyze the sales data at retailer/SKU level for a global
cosmetics brand (SAS, R).
– Ran a VAR Model to analyze the sales and customer mindset tracking data at DMA/Channel/SKU
level for a global beer brand (SAS, R).
– Built a classification tree for a segmentation study (CART, RandomForest, R).
• Verizon Data Services Tampa, FL
Technical Lead June 2001 – January 2011
– Development of provisioning, service assurance and surveillance software for Verizon’s FiOS
project (C++, Java, SQL).
– Providing decision support to business to improve customer experience (SQL, Perl).
– Pro-actively monitoring the network to drive down customer call volume.
• Primeon Burlington, MA
Senior Consultant July 1999 – April 2001
– Development of a gateway between Spectrum and CA Unicenter products.
– Providing professional consulting services of Spectrum network management platform.
– Providing professional consulting services of Interwoven Teamsite content management sys-
tem, mainly designing workflows and templates.
• Cattsoft Beijing, China
Senior Software Engineer April 1996 – July 1999
– Performed overall system design for a telecommunication management network conformance
testing system, sold to a Telecom operator in Beijing.
– Developing software for managing Telecom network.
Education
• University of South Florida Tampa, FL
Ph.D., Economics 2011
– Dissertation: Modeling Endogenous Treatment Effects with Heterogeneity: A Bayesian Nonparametric
Approach
– Major Professor: Dr. Murat Munkin
• University of South Florida Tampa, FL
MBA, Finance 2005
• Beijing University of Aeronautics & Astronautics (Bei Hang) Beijing, China
Master, Computer Science 1996
• Dalian University of Technology Dalian, China
BA, Naval Architecture 1993
Technical Skills
Programming Languages: C++, R, Java, Perl, Python, Shell scripts, SQL
Econometrics/Stat Lang/Packages: R, Matlab, SAS, Stata
Systems: Linux (CentOs/Fedora/Ubuntu), Oracle, DB2, MySQL
Libraries/Tools: Boost, MKL/ACML (BLAS/LAPACK), Armadillo, Ipopt, Vim, Git
Bayesian Statistics: MCMC, Gibbs, Metropolis-Hastings, Hamiltonian MH, Reversible Jump, Slice
Sampling
Statistical/Machine Learning Algorithms: Naive Bayes, CART, RandomForest, MARS, SVM, KNN
Statistical/Econometric Models: Finite Mixture Models, Dirichlet Process Mixture Models, Discrete
Choice Models, Hierarchical Bayes Models, Hidden Markov Models (HMM), Vector Autore-
gressive (VAR) models, Zero-Inflated Poisson (ZIP)