Analysis and Research on the Use Situation of Public Bicycles Based on Spark Machine Learning

Abstract

Public bicycles are a healthy and environmentally friendly means of transportation that facilitates people's travel. However, due to the uncertainty of urban travel, especially the tidal phenomenon, public bicycles often "difficult to borrow a car" and "return the car". This will result in unreasonable distribution of the site during the operation of the public bicycle system, unbalanced bicycle processes at various sites during peak hours, and unbalanced operation and management, which restricts the development of public bicycles. This paper uses the data of the San Francisco Bay Area as the experimental data of this paper, using Spark SQL and Spark Dataframe to analyze the use of public bicycle users and sites, according to the impact of different user types on the use of public bicycles, using K-means clustering algorithm Analyze the use of the site. Based on the Spark MLlib machine learning library, the gradient usage algorithm is used to predict daily usage.