Webscraping, correlation, and Data Analysis: Uncovering predictors of Netflix Paid Membership: Webscraped and structured data about Netflix Originals. Uncovered a significant predictor that impacts Paid Membership with extensive EDA to help hedge funds gain a competitive edge.
Topic: Kaggle Competiton - IEEE-CIS Fraud Detection - Project Presentation.
Models and techniques used: Gradient Boosting, SMOTE, Stratified Cross-Validation
Description: Utilized machine learning techniques such as SMOTE and GBM to improve the efficacy of fraudulent transaction alerts to empower businesses to reduce fraud loss and strengthen competitiveness in Kaggle.
Completion time: One week
Presentation Date: September 19, 2019 Location: NYC Data Science Academy Speakers: Fred(Lefan) Cheng, Luke Gray, Rajesh Earlu, Jose Gonzalez, Alyssa Wei
Code: https://github.com/LefanCheng My Linkedin: https://www.linkedin.com/in/lefancheng/ My Email: fredchengnyc@gmail.com