RentoBikes, a leading US bike-sharing provider, is addressing pandemic-induced revenue challenges by analyzing factors influencing bike demand. Engaging a consulting firm, they aim to predict demand post-lockdown and tailor their strategy. Data preparation involves converting numeric variables and retaining the 'yr' column for its predictive value. The model, focusing on the 'cnt' variable, aims to reveal insights into demand dynamics, guiding BoomBikes in optimizing their services for post-Covid market recovery
- Understanding and Reading the Data
- Data and Libraries
- Data Inspection
- Check for the correlation
- Insights Derivation
- Dealing with Categorical Data
Upon working with the data the insight that came into the picture were quite interesting
Insights
- Seasonal Analysis
- Fall has the highest average rentals, followed closely by summer
- Year-wise Rentals
- 2019 sees a significant increase with the median rise of approximately 2000 rentals compared to 2018
- Monthly Trend
- September tops the monthly rental count, with surrounding months showing an increase in demands
- Holiday vs Working Day
- Holiday generally shows that there are lower rental counts as compared to working days
- Holidays also exhibit greater variability in rental demands
- Weekday Analysis
- Overall, there is no significant difference in rental across weekdays is observed
- Thursdays and Sundays stand out with higher variability in rental counts as compared to other weekdays