Prediction and Analysis of Shared Bicycle Rental Demand Based on Feature Engineering and Regression Algorithm

This paper utilizes machine learning algorithms to analyze various types of data of shared bicycles and predict the rental demand. It conducts a correlation analysis based on the influence of factors such as time, holidays, weather, season, and humidity on the demand for shared bicycles. Feature data is screened according to the correlation coefficient and a feature dataset is constructed. By comparing the real values with the predicted values after model training, it is found that the random forest regression model has better prediction performance in this study.