Lester Tsai

Credit Modeler Intern

Conn's HomePlus•  June 2021 - September 2021

 Built Regression and Tree-based machine learning models in Python to decide on client loan approval utilizing hyperparameters tuning and cross-validation to reach an accuracy of over 0.7 on average.  Accelerated computation speed by building pipelines in PySpark, saving 10% time compared to Pandas.  Identified fraud applications with over 80% correctness from application information on the SentiLink platform.  Created and maintained monthly credit Excel reports using Visual Basic for Applications (VBA), including visualizing different metrics and KPIs and updating pivot tables.  Assisted the managers in selecting the best model before deployment for the business need.

Statistical Assistant

Greenpeace•  June 2021 - August 2021

 Web-scraped 1000+ pieces of news from 5 websites to gather all the environmental-related news (50) in the past month to locate polluted regions in Taiwan utilizing Requests and BeautifulSoup packages in Python.  Trained and supported the researchers in efficiently collecting and wrangling data using Pandas and BeautifulSoup instead of manually using the spreadsheet.

University of Washington, Washington

Data Science, MS•  September 2021 - Present

Rigorous training in advanced statistical theorem and its application to public health.

National Taipei University

Mathematics, BS•  September 2017 - June 2021

Rigorous training in fundamental statistical theorems and the capability of using software to conduct data-analysis projects, including Python, R, and SAS.