Projects
One Billion NYC Taxi Trips Analytics in Hive
An open-source exploration of NYC's neighborhoods, nightlife, airport traffic, and more.
Designed data warehouse schema in compliance with the star schema architecture.
Sliced and diced data based on date dimensions and business timings. Created dimension and fact tables, took data scan cost, query performance and storage cost into consideration, loaded data with Hive scripts.
Implemented UDFs with Python to add helper attributes. Summarized the pick-up and drop-off patterns, traffic flow trends and heat maps of trip connections aggregated over space in the final report.
View Project
NYC Zip Code Investment Analytics
Formalized the business problem and identified the key factors of real estate market price. Built simplified profit prediction model based on property market price and rental revenue.
Retrieved historic real estate sales prices from Zillow APIs and home rates and availability from Airbnb.
Performed thorough data cleaning, involving insufficient records elimination, estimation of missing prices and occupation rate, etc. Plugged data into model and calculated ROI.
Visualized metrics for profitability on short term rentals by zip code, summarized our key insights into future real estate investments with seaborn and matplotlib.
View Project