Bike Sharing Assignment

BoomBikes, A US basedbike-sharing provider has recently suffered considerable dips in their revenues due to the ongoing Corona pandemic. The company aspires to understand the demand for shared bikes among the people in the American market after this ongoing quarantine situation ends across the nation due to Covid-19.

The company wants to know:

1. Which variables are significant in predicting the demand for shared bikes.
2. How well those variables describe the bike demands

Table of Contents

• General Info
• Technologies Used
• Conclusions
• Acknowledgements

General Information

• Provide general information about your project here. This project is aimed at finding the variables significant in predicting the demand for the shared bikes thereby helping the company to avoid loss.

• What is the background of your project? This is an assignment study taken as a part of AI&ML Jan 2024 batch IIIT-Bangalore

• What is the business probem that your project is trying to solve? Identifying the variables that help in predicting the demand for shared bikes

• What is the dataset that is being used? day.csv - This file contains data spanning two years (2018-2019), detailing the number of users renting bikes per day. Each day is characterized by various attributes such as working day, temperature, season and more.

Conclusions

• The r2 is 80.4 on test set and on trian set it is 83.3 with lr_model. The difference is less than 5%
• Demand for the bikes depends on following independent features: temp, holiday, windspeed, year, September, Spring, Summer, Winter, Mist, weathersit -3 (Light Snow, Light Rain + Thunderstorm + Scattered clouds, Light Rain + Scattered clouds) as per built model lr_model_rfe
• Further from EDA analysis:
• Company should focus on expanding buisness in September
• Company can provide offers during weekends/holidays as the users count is high these days
• During Adverse weather conditions user count is too low, company can focus on vechile maintainence.
• Year by year the user count has increased, so this might give a clue that post pandamic situation can improve. So, company should be prepared to handle user needs accordingly.

Technologies Used

• pandas library - version 2.0.3
• numpy library - version 1.24.3
• matplotlib library - version 3.7.2
• seaborn library - version 0.12.2
• statsmodels library - version 0.14.0
• sklearn library - version 1.3.0

Acknowledgements

Give credit here.

• This Linear Regression assignment was taken as a part of AIML - Jan 2024 batch IIIT-Bangalore

Contact

Created by [@smarunkumar] - feel free to contact me!