Jiaman Betty Wu

Jiaman Betty Wu

Data Scientist

About me

Hello! My name is Jiaman Wu (吴佳蔓). My friends usually call me Betty.

I studied economics and statistics in college. Although I enjoy a conversation about investing or a friendly debate on frequentist vs Bayesian statistics, I eventually landed a career in data science. I recently finished my Master’s degree. And I’m currently working as a marketing data scientist. My work focuses on market acquisition through the thoughtful application of experimental designs and machine learning models.

Education
  • M.S. in Data Science, 2020 - 2022

    Duke University

  • B.A. in Statistics, 2015 - 2019

    University of California, Davis

  • B.A. in Economics, 2015 - 2019

    University of California, Davis

Experience

 
 
 
 
 
Capital One
Data Scientist
Jun 2022 – Present McLean, VA
  • Develop and maintain marketing models that score customers’ conversion propensity on a bi-monthly basis.
  • Manage model risks through active model monitoring. Address potential model errors or drifts.
  • Contribute to building a culture of experimentation by advising business teams on A/B testing best practices. Collaborate with business partners to refine experimental designs. Build tools to enable appropriate statistical tests and power analysis for stakeholders.
 
 
 
 
 
Duke University Marine Lab
Data Scientist
Duke University Marine Lab
Aug 2021 – Apr 2022 Durham, NC
  • Partnered across multiple institutions to deliver a machine learning pipeline for estimating the abundance of endangered North Pacific whales with acoustic recordings from the sub-Arctic.
  • Improved and automated the existing abundance estimation process by developing CNN models that extract whales’ “gunshot” calls.
 
 
 
 
 
Duke University School of Medicine
Student Researcher
Duke University School of Medicine
Aug 2021 – Apr 2022 Durham, NC
  • Partnered with the Department of Anesthesiology to deliver a machine learning solution to estimate the expected opioid dosages for various medical procedures.
  • Identified abnormal opioid dosages and flagged at-risk healthcare providers using the resulting model.
  • Led the effort in identifying and removing outliers in multi-year medical records in collaboration with medical professionals. Redesigned the feature engineering process to reduce data dimension. Improved model r-squared by 30%.
 
 
 
 
 
Duke Graduate School
Teaching Assistant
Aug 2021 – Dec 2022 Durham, NC
  • Mentored graduate students' projects in applied statistical inference and statistical learning.
  • Held weekly office hours. Responded to student questions on course forums. Graded assignments and projects.
 
 
 
 
 
Capital One
Data Science Intern
Jun 2021 – Aug 2021 McLean, VA
  • Led the effort of defining and measuring gaming behaviors in incentive marketing campaigns.
  • Collaborated with cross-functional teams to gather information and ensure data integrity in the collection process and downstream analysis.
  • Prototyped propensity models for identifying potential gamers using Databricks, Snowflake, and Scikit-Learn.
  • Communicated insights to business partners and stakeholders by incorporating principles of data storytelling and interpretable machine learning.
 
 
 
 
 
Davis Sensory Institute
Principal Data Analyst
Aug 2020 – Apr 2021 Davis, CA

As a member of an early-stage start-up, I actively involved in all aspects of the business and helped to estalish common practices and guidelines in data analytics. My main efforts include:

  • Initiated the movement from Excel to R, which significantly expedited the data analysis process and the rotation table designing process.

  • Discovered discrepancies and errors in the consumer study database, cleaned the existing database, and enhanced the data collection process to ensure accuracy and efficiency.

  • Analyzed product data and extracted product insights by employing inferential statistical testing and data visualization. Innovated the analysis for text data by introducing sentiment analysis in Natural Language Processing.

  • Led report writing to communicate findings and business recommendations to clients, including Fortune 500 companies. Delivered workshops regularly to train interns in adequate statistical and programming skills.

Projects

*
Healthcare Transportation Challenge Predictive Modeling

Healthcare Transportation Challenge Predictive Modeling

Trained and validated classifiers to identify high risk patients with transportation challenges. Advanced as 2nd round finalists out of 279 competing teams.

Poisonous Mushroom Classification

Poisonous Mushroom Classification

Built and evaluated a logistics regression model to seperate edible and poinous mushrooms by engineering features and performing rigorious model selections. Identified and interpreted important features.

Serverless Language Translation Flask Web App

Serverless Language Translation Flask Web App

Built a serverless continuous delivery pipeline for language translation on AWS. Launched the web app that is capable of provisioning, deploying and scaling automatically.

The Effects of Ressions on Educational Attainment

The Effects of Ressions on Educational Attainment

Built econometric models to identify relationships between economic recessions and educational attainment.

Skills

Machine Learning
Programming

Python, R, SQL

Cloud Technology

AWS

Accomplish­ments