Python R Java SQL Machine Learning Modeling Visualization

Bangda Sun

******@********.*** • 917-***-****

https://www.linkedin.com/in/bangda-sun

EDUCATION

COLUMBIAUNIVERSITY

GRADUATE SCHOOL OF ARTS AND

SCIENCES

M.A. IN STATISTICS

Expected Dec 2017 New York, NY

Cum. GPA (1st Semester): 4.00

ZHEJIANG GONGSHANG UNI-

VERSITY

SCHOOL OF STATISTICAL AND

MATHEMATICS

B.A. IN STATISTICS

Grad. Jun 2016 Hangzhou, China

Cum.GPA: 3.79 Major.GPA: 3.92

Rank: 1/93

COURSEWORK

GRADUATE

Probability

Statistical Inference

Linear RegressionModel

Statistical Computing and Intro toData

Science

Statistical Machine Learning

Statistical Methods in Finance

Data Structures in Java

UNDERGRADUATE

DataMining

Time Series Analysis

Multivariate Statistical Analysis

Econometrics

Sample Survey

Intro toDatabase

Mathematical Analysis (I) and (II)

Linear Algebra

Intro to Statistics

Intro to Actuarial Science

PROGRAMMING SKILLS

R • Python •Matlab • SQL•

Java • SAS •HTML • LATEX

INTERNSHIP EXPERIENCE

YIWUWHOLESALEMARKET DATA ANALYST IN INDEX OFFICE July 2015 – Sep 2015 Jinhua, Zhejiang, China

• Sampled the product sales from10,000+ booths in toy industry two times every week, and checked the data collected by other collectors

• Involved in Yiwu-Index computing and publication: used the collected data to compute gross turnover, online sales, etc, and imported the data into five databases (used SQL)

• Submittedweekly technical report based on the trend of Yiwu-Index, included the statistical summary of themarket operation, analysis and prediction of the sales based on Yiwu-Index

PROJECT

REGRESSIONANALYSISONHOUSINGPRICE INBAYAREA

(DATASET FROMKAGGLE) Dec 2016 New York, NY

• Used scatter and box plot to do variable selection among 18 variables to build a basic linear model

• Used best-subset method and LOOCV to find the best predictors, comparing models byMSE on testing set

• Analyzed the residuals of models, improving themodel based on Bias-Variance trade-off principle by doing logarithm transformation, adding dummy variables, adding higher termpredictors and interactions, increasing the prediction accuracy by 15%

EXPLORATIONONCONSTRAINTSOFWECHAT BUSINESS

DEVELOPMENT Apr 2015 - Nov 2015 Hangzhou, Zhejiang, China

• Designed Likert scale tomeasure the influence level fromkey factors and applied Stratified and Systematic Sampling and collected 300wechat business cases online and offline.

• Used Factor analysis model andmultivariate linear regressionmodel by R to quantify the relationship between factors which influence the development of wechat business

PUBLICATION

ANANALYSISONTHE INFLUENCEOF AIRQUALITYON

RESPIRATORY SYSTEMDISEASE INCIDENCE INMININGAREA SHANGHAI ENVIRONMENTAL SCIENCES 2015: VOL.34 NO.02, 67-69 Apr 2015 Hangzhou, Zhejiang, China

AWARDS AND INTERESTS

2015 National Scholarship top 38 in Zhejiang Gongshang University 2015 Meritorious winner inMCM top 9%

2014 1st Prize Award in CUMCM top 1.3%

Contact this candidate