Zheyan Liu

EDUCATION

Columbia University - Mailman School of Public Health Expected May 2023

Master of Science in Biostatistics, Specialization in Data Science
Relevant Coursework: Introduction to Database, Data Science (R), Biostatistics Method

Shanghai University of Finance and Economics (SUFE) - School of Statistics and Management, China June 2021

Bachelor of Science Statistics
Relevant Coursework: Data Mining (Python), Machine Learning, Computational Programming (C++), Mathematical Analysis, Probability, Advanced Mathematics, Regression Analysis (R, SAS), Random Process, Database Theory (SQL, C#), EDA (R)

PROFESSIONAL EXPERIENCE

Ping An Technology | Data Scientist Intern , Shanghai, China April 2021 – July 2021

• Implemented and deployed of data pipelines, data models, and ETL processes with PySpark in Apache Zeppelin
• Optimized web crawlers code and feature engineering process; improved efficiency by 40%
• Used LightGBM, XGBoost to predict if customer will renew auto insurance and health insurance with AUC 0.93; researched the feasibility of implementing the GNN auto insurance anti-fraud model

Shengqu Games | Data Analyst Intern , Shanghai, China October 2020 – April 2021

• Reduced reporting time from hours to seconds by using SQL and Python to query data and automate reporting
• Researched on user payment/churn issues and provide in-depth data reports to the game planner teams with Tableau and Plotly
• Predicted user churn with AUC 0.92 using Kmeans, RFM analysis, and an ensembled model of XGBoost and Random Forest

Hylink Digital Solution | Data Analyst Intern, Shanghai, China August 2020 – October 2020

• Wrote Python code to update daily/weekly report and built a new data dashboard on Excel to demonstrate performance of campaign to our client (PayPal) with convenience
• Undertake exploratory data analysis (using Python and Tableau) towards weekly campaign data on several media such as Baidu and Google to offer business insights to our client

PROJECT EXPERIENCE

New York Times Best-seller Book Pipeline| Team Leader, New York, United States Feb 2022

• Created DAG in Airflow to extract data from NYTimes Books API and sent the results to customized email recipients weekly.
• Transformed and cleaned the format of the extracted data and containerized the project with Docker

Azure Data Factory Project on Covid19 Reporting| Team Leader, New York, United States Dec 2021 – Jan 2022

• Created Data Flows to ingest and transform data from sources as HTTP; Scheduled pipelines with triggers to update daily
• Leveraged Azure Monitor and Log Analytics to monitor pipelines and used Power BI to build a reporting dashboard

NYC Subway Crime GNN Prediction and Application | Team Leader, New York, United States Nov 2021 – Dec 2021

• Cleaned and imputed 10 years of New York City subway passenger data; Built project website and organized report
• Conducted EDA on subway passenger flow and location; Clustering station coordinates into 8 clusters with K-means
• Built Crime-free Subway Route Identification Rshiny App based on Google Map Api and adapted GNN to predict crime in each route

Sentiment Analysis of Covid19 Internet Public Opinions | (Graduation Thesis), Shanghai, China April 2021 – May 2021

• Performed exploratory data analysis based on the text data related to the epidemic situation in Weibo.
• Used methods such as back translation, Bert synonymous replacement, and random replacement for data strengthen.
• Establish and tune BERT and FastText-based SVM and XGBoost models to predict sentiment tendency with AUC 0.89
• Applied voting model to the crawled epidemic-related Weibo data, and analyzed online public opinion of Covid

Empirical Covid19 and market volatility | Research Assistant, Shanghai, China April 2020 – June 2021

• Determined value for each parameter in SEIRD model through reading plenty of related papers and found the underlying relationship within variables
• Programmed using Python to simulate data under seven different circumstances
• Recorded findings and optimized scripts to simulate data efficiently and increase detailed markdown and annotation
• Conducted sensitivity test on parameters to give Covid-control policy suggestions

Design of Database for SUFE Volunteer Management System | Team Leader, Shanghai, China Nov 2019 – Jan 2020

• Co-analyzed needs of project, drawn E-R diagram to optimize data model and used SQL Server to establish database
• Developed user platform with C# Windows Forms to interact with database and responsible for testing and debug