• Enhance collection efficiency by conducting data analysis to understand customers’ profiles and to perform data segmentation• Champion continuous process improvements, upgrade collection tools and optimize collections systems• Lead daily and monthly collection reporting on a timely basis with high level of accuracy• Developed a data-driven approach to guide management’s decision-making on collection strategy• Increased money collected by using credit score to identify high risk customers for early action, through an automated process in Python

• Delivered complex concepts in a way that non-technical audiences can understand and be able to convince non-technical/technical partners with well-explained process and methods• Ensured data and machine learning production pipelines are scalable, repeatable, and secure. Automated data pipelines to make data reporting and data engineering process more efficient• Implemented surveys to collect data and developed dashboards to visualize data for policy-makers to make evidence-based decisions• Organized the 2022 International Methodology Symposium by deriving the scientific aspect of the conference and discussing topics for invited sessions and workshops• Built XGBoost models in R to predict the Retail Services Price Index with less than 5% prediction error in real-time, and improved the timeliness of releasing official statistics• Outperformed the ARIMA forecast model by customizing the XGBoost model learning process and tuning the model parameters using the rolling window forward validation protocol• Evaluated performance of different models by visualizing the nowcasting trend and prediction error• Utilized multivariate allocation to minimize sample sizes and collection costs for publishing accurate statistics• Improved data quality and fitness to use by imputing missing values with linear regression models and nearest neighbour approach in SAS• Reduced nonrespondents by understanding nonrespondents’ characteristics and behaviours using k-medians clustering algorithm• Performed data extraction and manipulation over large relational data sets via SQL joins• Identified imputation class variables using a multinomial logistic regression model in SAS

• Ensured electronic medical records are of high quality by dealing with missing values and removing data discrepancy in spreadsheets• Quantified the effect of transthoracic echocardiography on patients' mortality using a weighted logistic regression model in R• Visualized research findings in boxplots and prepared a manuscript to interpret the results in a statistical framework

• Presented statistical concepts and modeling procedure using interactive activities and PowerPoint in biweekly SAS tutorial sessions• Collaborated with a team of 5 teaching assistants and 2 professors each semester to perform various duties including proctoring exams and marking students' assignments• Assessed students' project proposals and provided constructive, written feedback to enhance students' critical thinking ability and to improve future performance

• Utilized financial knowledge about option pricing and hedging strategies to tutor students • Held monthly tutorial sessions to teach 30 students how to conduct regression analysis and time series forecasting in Excel