Curriculum Vitae

Education

Candidate for B.S. (Hon.) in Mathematics and Applied Mathematics, Zhejiang University, China

Date: Sept. 2014–Present
Member of the Qiushi Pursuit Science Class (Mathematics), an exclusive program for intensive training in mathematics
Overall GPA 3.83/4.0 (third-year GPA 3.99/4.0), GRE 332+4.0 (V 162, Q 170), TOEFL 111 (Speaking 26)
Mathematics Coursework (selected): Abstract Algebra, Mathematical Analysis(A), Probability Theory, Stochastic Processes, Ordinary Differential Equation(A), Point Set Topology, Real Variable Analysis, Differential Geometry
Computer Science Coursework (selected): Lectures on C Programming, Fundamentals of Data Structure, Advanced Data, Structure and Algorithm, Object-Oriented Programming, Database Technology, Artificial Intelligence

Visiting Undergraduate Researcher, in the Dartmouth Machine Learning Group, Department of Computer Science, Dartmouth College, NH

Date: Jul. 2017–Oct. 2017

Exchange Student, Department of Mathematics, Ohio State University, OH

Date: Aug. 2016–Dec. 2016
Term GPA 4.0/4.0
Coursework: Honors Combinatorics, Intermediate Data Analysis, Applications of Differential Equations

Structured Stein Variational Inference for Continuous Graphical Models

D. Wang∗, Z. Zeng∗, and Q. Liu. Structured Stein Variational Inference for Continuous Graphical Models. ArXiv e-prints, November 2017. (∗Equal contribution, will submit to ICML 18)

Research Experience

Project: Subset selection in probabilistic graphical models using Stein lower bound

Advisor: Prof. Qiang Liu, Department of Computer Science, Dartmouth College
Date: Oct. 2016–Present
Proposed Stein lower bound for estimating Bayesian optimal risk and analyzed its finite-sample confidence bound
Carried out subset selection experiments on probabilistic graphical models and affirmed that in some cases, the Stein lower bound gave tight estimation and was effective in selection tasks compared with baseline methods.
Currently try out more applications and put the theoretical analysis and experimental results into a paper

Project: Variational inference algorithm for probabilistic graphical models

Advisor: Prof. Qiang Liu, Department of Computer Science, Dartmouth College
Date: Jul. 2017–Nov. 2017
Proposed the Graphical Stein variational gradient descent algorithm, a distributed inference algorithm for continuous graphical models as an extension of Stein Variational gradient descent (SVGD) by leveraging the Markov dependency structure
Analyzed the theoretical validity of the algorithm by studying Stein discrepancy and also analyzed convergence properties
Applied the proposed algorithm to applications including crowdsourcing and location estimation of sensor network
finished the paper-writing on this work and posted the paper to arXiv as joint first author

Project: Modeling the olfactory learning of honeybees with spiking neural networks

Advisor: Prof. Nenggan Zheng, Advanced Computing and System Lab, Zhejiang University
Date: Apr. 2016–Jun. 2017
Studied the biological topology of lateral antennal lobe tract (l-ALT) in honeybee olfactory system and hypotheses on the links between individual learning and changes in structural organization in nervous system
Developed a model for l-ALT using spiking neural networks and optimized the STDP learning rules for computational efficiency
Analyze the plausibility of hypotheses by comparing the biological recordings of the circuit and model output

Zhe Zeng