Skip to content

Latest commit

 

History

History
16 lines (12 loc) · 858 Bytes

README.md

File metadata and controls

16 lines (12 loc) · 858 Bytes

HaggisHopper

Leveraging K-Means clustering and advanced visual analysis and ML forecasting models to gather business insights for fleet optimization, enhance operational efficiency, design targeted customer-centric promotions.

Data Preprocessing

  • Cleaned and prepared trip data.
  • Handled outliers and engineered features like revenue per kilometer and average trip duration.

Exploratory Data Analysis

  • Visualized demand trends, revenue zones, and time-based patterns using heatmaps and Tableau.

Machine Learning

  • Applied K-Means clustering for hotspot identification and trip segmentation.
  • Built LSTM and Holt-Winters models for demand forecasting.

Geospatial Insights

  • Used geospatial visualization tools like Folium and Geopandas to map revenue and demand density by postcode.