Bike Rental Demand Forecasting Dashboard

Project Overview-

This project focuses on predicting hourly bike rental demand to ensure a stable supply in urban mobility systems . By analyzing historical data, weather patterns, and temporal factors, the dashboard provides actionable insights for fleet rebalancing and accessibility optimization.

Key Features-

Predictive Analytics: Forecasting total rentals for the next 24 hours based on historical trends.

Station-Level Insights: Identifying high-demand clusters and rebalancing needs using geospatial heatmaps.

Weather Impact Analysis: Visualizing how temperature, humidity, and rainfall affect rental behavior.

Scenario Simulation: Interactive filters to simulate demand under different weather and holiday conditions .

Technology Stack-

Analytics Platform: Microsoft Power BI.

Data Engine: Power Query for ETL processes.

Forecasting Method: Built-in Time-Series Forecasting (Exponential Smoothing).

Metrics: Developed using DAX (Data Analysis Expressions).

Key Performance Indicators (KPIs)

Predicted vs. Actual Accuracy: Consistently high R-Squared scores (e.g., 0.88).

Error Margin: Mean Absolute Error (MAE) maintained within optimal limits (e.g., 1.2 km).

Peak Demand Mapping: Identifying peak hours (e.g., 6:00 PM) for better resource allocation.

Project Structure-

/dashboards: Contains high-resolution screenshots of the Power BI interface.

SANKALP SOHGAURA - MS Elevate AICTE Internship - Project.pdf: Detailed final presentation .