allisongao4015 / ds-bayesianism_bvh32 Goto Github PK

I am a policy school graduate and Flatiron Data Science Bootcamp grad with a passion for data, tech, and policy analysis. I am currently working at the New York City Taxi and Limousine Commission, working at the intersection of data analytics and policy research to support the agency's policy initiatives for the for-hire transportation sector.

• Classify whether an issued civil ticket in NYC will result in fine collection using machine learning
• Analyzed over 210,000 civil tickets issued in New York City using 25 associated variables
• Automated download of over 200 files from the Census Bureau using Selenium
• Ran iterations of logistic regression, random forest, decision tree, and XGBoost models to select best model
• Selected random forest as the preferred model with 71% accuracy score and 66% precision score

• Classify Tweets on Google and Apple products into positive emotion, negative emotion or no emotion detected
• Trained 4 different machine learning classification models
• Used different text vectorization tools and machine learning tools to fix class imbalance
• Identified Multinomial Bayes as the best model with a 77% macro precision score

• Multiclass classification modeling to predict risk of injury resulting from traffic crashes in Chicago
• Merged 3 datasets with over 550,000 observations for analysis
• Ran multiple machine learning models and used grid search to identify best parameters
• Selected XGBoost as the preferred model with 95% accuracy score and 87% in macro precision score

• Linear regression modeling to analyze a myriad of housing factors’ impact on house price in King County
• Used home sale data on 15,000 houses and 23 variables to conduct linear regression modeling
• Analyzed and interpreted variables’ coefficients to identify important features associated with house price