Analysis of taxi track data based on spark platform

How to analyze a large amount of traffic data, improve the level of urban traffic management, and discover the travel rules of urban residents from complex data, and rationally dispatch vehicles. Combined with the trajectory data of the yellow and green taxis in New York City, the traffic data is processed based on the Spark distributed data processing platform, and the K-means clustering algorithm is used to analyze the passengers' access points. The visualization of taxi passenger travel characteristics based on the platform is given, including the impact of rental car passengers, the distribution of urban residents' travel time and the distribution of taxi area speed. The gradient lifting algorithm is used to predict the importance of each feature of the training set, so as to rationally dispatch the vehicle.