Yifan Wu
About Yifan Wu
Yifan Wu is a Data Scientist specializing in machine learning for personalization at Credit Karma, where he has worked since 2020. He holds dual degrees in Mathematics and Applied Statistics from Penn State University and a Master's in Financial Mathematics from The University of Chicago.
Work at Credit Karma
Yifan Wu currently serves as a Data Scientist specializing in machine learning for personalization at Credit Karma. He has been with the company since 2020, contributing to the development of optimization strategies aimed at enhancing user engagement. His role involves leveraging machine learning techniques to create personalized experiences for users, which is critical for the company's mission to empower consumers with financial insights.
Education and Expertise
Yifan Wu holds a Bachelor of Science (BS) in Mathematics with an Actuarial Science Option and Applied Statistics, along with a Minor in Economics from Penn State University, where he studied from 2013 to 2016. He furthered his education by obtaining a Master of Science (M.S.) in Financial Mathematics from The University of Chicago, completing his studies there from 2016 to 2018. His academic background provides a strong foundation in quantitative analysis and statistical modeling.
Background
Yifan Wu's professional journey includes diverse experiences in data science and analytics. He began his career as a Summer Analyst at Accenture in 2014 in Beijing, China, followed by a Summer Actuarial Analyst position at American International Group (AIG) in 2015 in Houston, Texas. He worked as a Risk Modeling Intern at Northern Trust Corporation in 2017, gaining insights into financial risk assessment. Prior to joining Credit Karma, he spent two years as a Data Scientist at ADARA, Inc. in the San Francisco Bay Area.
Achievements
During his tenure at Credit Karma, Yifan Wu has designed an innovative modeling framework utilizing Semi-Supervised learning to tackle class imbalance issues in opt-out models. He has also implemented model calibration approaches on Click-through Rate (CtR) to improve the responsiveness of machine learning models to market changes. Additionally, he created a periodic and automated data preparation workflow using Airflow, which significantly streamlined the model refreshing process.