Dong Ming Zhen Email & Phone Number

New York City Metropolitan Area

Completed an immersive accredited data science bootcamp that prepares students with the necessary skills to successfully apply machine learning and data analysis to real world data problems. Projects require a strong understanding of Python programming, statistics, and project design skills as well as the ability to communicate the results to stakeholders. Projects include:Find fishing locations and gear used by local NY fishermen - Designed a data pipeline that scraped… Show more Completed an immersive accredited data science bootcamp that prepares students with the necessary skills to successfully apply machine learning and data analysis to real world data problems. Projects require a strong understanding of Python programming, statistics, and project design skills as well as the ability to communicate the results to stakeholders. Projects include:Find fishing locations and gear used by local NY fishermen - Designed a data pipeline that scraped, cleaned, and merged 30,000 posts from NyBass and NyAngler- Used spaCy and RegEx to remove stop words and lemmatize the text data- Created a documented term matrix with a TDIF vectorizer and applied NMF to find relevant fishing topics from reoccurring terms- Extracted strongly defined topics to make fishing location and gear recommendations by seasonPredict Delta Airline's departing flight delays- Constructed a data pipeline to automate data collection and cleaning for the top 25 airports in the U.S. for Delta Airlines for a specific year- Chunked a IBM airline dataset with 200 million rows to find the relevant rows- Invented new features to help with prediction such as the scheduled number of flights a plane has in the day- Created a stratified cross validation pipeline to address the class imbalance and trained the data on logistic regression, random forest, and xgBoost modelsPredict NBA player's Fanduel score- Scraped NBA team statistics, roster, and player box score data from the 2017 - 2019 seasons from basketball reference with BeautifulSoup- Identified offensive and defensive tendencies and player attributes as the strongest predictors of a NBA player's Fanduel score; created features from existing data to proxy for missing player attribute data- Normalized skewed features using log transformation, and adjusted features to a polynomial of degree 2 to train a linear regression model across a five fold cross validation that made predictions with a mean absolute error of 4.9 Show less

Bachelor Of Business Administration (Bba), Marketing/Marketing Management, General

Education record