A forecast for bicycle rental demand based on random forests and multiple linear regression

Bike sharing system is a ways of renting bicycles; bike return is automated via a network of kiosk locations throughout a city. Using these systems, people are able to rent a bike from a one pick up location and combine with their as-need, customer returns bike to the place, which they would prefer to return. This paper is asked to combine historical usage patterns with weather data in order to forecast bike rental demand in the Capital Bike share program in Washington, D.C. Firstly, the multiple linear regression model was established by the conventional method, Multiple linear regression equation was obtained by using SPSS software, After comparing the data with the real value, it is indicated that the multiple linear regression model is less accurate. After analysis, we find that the data includes the dummy variables such as the time and the season. Hence this paper proposes a random forest model and a GBM packet to improve the decision tree. The results and the accuracy of multiple regression analysis are greatly improved when use of random forest model to predict the demand for bicycle rental.