Data analytics, Machine Learning, Python, R, SQL, Tableau

Jie Zhang

213-***-**** ********@***.*** www.linkedin.com/in/jie-jeffrey-zhang-04383615a Los Angeles, CA EDUCATION

University of Southern California (USC) Los Angeles, USA Master of Science in Analytics, GPA:3.9/4.0 Aug. 2019-May. 2021 Hefei University of Technology (HFUT) Hefei, China Bachelor of Economics, GPA:3.8/4.0 Sept. 2015-Jun. 2019 PROFESSIONAL SKILLS

• Technical Skills: Python, R, Tableau, MySQL, Advanced MS Office, Google Suites, SPSS, Stata, AMPL

• Data Science: Predictive Analysis, Time Series Analysis, Hypothesis Testing (A/B testing), Data Visualization

• Certification: Data Science (Coursera), Tableau Essential Training (LinkedIn) INTERNSHIP EXPERIENCES

Product Operation Intern Hefei, China

iFLYTEK - AI Marketing Group Apr. 2019-Jun. 2019

• Discussed with executive to identify the specific target number of ad requests and clicks in “JD 618 big promotion”

• Created dashboards to visualize daily and weekly CTR, exposure rate, CPC and CPT etc. of segments by Tableau

• Cluster advertisement (Focus Picture, News Feed, Screen Opening) based on provinces and advertisement types

• Made a prediction on indicators mentioned above to monitor abnormal value by R

• Aggregated statistical data of ad cost and ad performance for different clients, then created PowerPoint slides to report

• Increased daily average ad requests by 220% and daily average ad clicks by 187% year-on-year Credit Analyst Intern Rizhao, China

Agricultural Bank of China - Corporate Credit Team Jan. 2019-Mar. 2019

• Formalized and analyzed financial documents of 20+ business loan applicant companies to assess their asset liquidity, credit soundness, and operational cash flow

• Tracked and updated post-loan financial performance and monitored risk profile for 5 companies

• Obtained 10 years financial statements from client companies and used the bank’s proprietary credit loan model by Python to calculate default risks and finished clients credit rating reports PROJECT EXPERIENCES

Southern California Edison (SCE) Load Forecasting Project- Los Angeles Jan. 2020 - Feb. 2020

• Processed 40000+ real electricity load data records in southern California which come in hour intervals for everyday starting from 2014 to 2019 by adding categorical variables related to weekdays, months, holidays and temperature etc.

• Conducted Time-Series Analysis to fit data by building ANN and LSTM RNN model with Python, then diagnosed models’ performance by adapting several methods (MSE, AIC, R2)

• Found optimal hidden units in each hidden layer and time steps by grid search in LSTM model and adjusted 3 variables to 67 variables which reached a goal of a MAPE under 2% for 16-hour forecast TalkingData AdTracking Fraud Detection (Kaggle Competition in 2018) - Los Angeles Nov. 2019 - Dec. 2019

• Cleaned 180M user click records provided by Kaggle and selected 10 most related features by feature engineering

• Employed Grid Search Cross Validation with Python to fit the data and found the best parameters in classification machine learning models (XGBoost Classifier Model, Logistic Regression Model, Random Forest Classifier Model)

• Forecasted test data by XGBoost Classifier Model and reached an accuracy rate larger than 80% Determinants of Bone Density (Empirical Research) - Qingdao Dec. 2019 – Jan. 2020

• Acquired information of 9000+ patients from hospitals in Qingdao and built a dataset including 75 features by Excel

• Conducted reliability analysis to choose variables whose correlation value with bone density are larger than 0.1

• Implemented Variable Selection Algorithm to filter out 17 features that have statistical significance with multiple linear regression model according to AIC, and obtained the final regression model whose R2 is 0.27 after 8 iterations by R

• Created a bone density prediction model by Random Forest Algorithm whose explanation degree is 37.08% Analysis of Energy Efficiency and Influencing Factors - Hefei Aug. 2018 - May. 2019

• Utilized Data Envelopment Analysis (DEA) to calculate the energy trade efficiency of 58 countries from 2008 to 2017

• Analyzed the difference of energy trade efficiency when individual energy products or aggregated energy products are the dependent variable in extended gravity model

• Constructed a fixed-effects model with Stata to analyze the influence density and trend of traditional influencing factors

(GDP, Distance, Tariff rate, etc.), and that of China facilities interaction policies to energy trade efficiency

• Clustered countries into three segments according to national income level to build same models for robustness test

Contact this candidate