team 26

deepthi meghana

#Overview Analysis of Uber pickup data in New York City Dataset includes millions of trips (2014–2015) Goal: understand temporal patterns (time) and spatial patterns (location) of demand #Dataset Source: NYC Taxi & Limousine Commission (via FOIL request) Key fields: Date/Time of pickup Latitude & Longitude Base company code Size: ~4.5M trips (2014) ~14M+ trips (2015) Data Preparation Converted Date/Time →: Hour Day Month #Extracted features: Weekday vs Weekend Time-of-day bins (morning, evening, night) Cleaned missing or invalid coordinates Exploratory Analysis 1. Pickups by Hour Peak demand during: Evening (5 PM – 8 PM) Late night (weekends) Lowest demand: Early morning (2 AM – 6 AM)

Insight: Uber is heavily used for commuting and nightlife

  1. Pickups by Day Highest: Friday & Saturday Lowest: Sunday / early weekdays

Insight: Strong weekend effect

  1. Monthly Trend Increasing pickups from April → September Suggests: Growth in adoption Seasonal effects (summer activity) #Spatial Analysis Highest pickup density in: Manhattan (central business districts) Moderate: Brooklyn & Queens Lowest: Staten Island

Insight:

Demand correlates with population density + commercial activity #Clustering Insights Using clustering (e.g., K-means): Identifies: Business zones Residential areas Entertainment hotspots

Insight:

Different zones show distinct time-based behavior #Key Findings Demand is not uniform (varies by time & location) Uber dominates: Evenings + weekends High-density urban zones Growth trend shows increasing reliance on ride-hailing Implications Useful for: Demand prediction models Traffic planning Ride allocation optimization Data can support urban transport policy decisions Conclusion Uber usage reflects: Human mobility patterns Work + social behavior Combining time + location analysis gives powerful insights into city dynamics