Abstract
The bike-sharing system allows people to rent a bike at one of the rental stations, use them for their journey, and return them to any other or the same station. In recent years, such systems encourage people to use bikes as a good complement to travel using modes of transport. The main purpose of this article is to predict the demand for shared bikes using multiple linear regression. Moreover, the article presents data cleaning process and data visualization for better understanding and to get useful insights from data. The results we get after visualization is that count of bike sharing is least for spring, the number of bikes increased in year 2019 and count of total rental bikes including both casual and registered increases in summer and are less in holidays. To understand which attributes needs to get dropped we used RFE (Recursive Feature Elimination) and VIF (Variance Inflation Factor) to drop columns with high p-values and to check the multicollinearity respectively. Now the model is able to predict the future demand for shared bikes in a city.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have