Analysis And Mining Of Traffic Hotspots In New York City Based On Taxi Passenger Data

Taxi as an important travel tool,is favored by more and more travelers because of its wide coverage,fast speed,comfort and safety.With the enrichment of people’s social activities and the development of big data,Internet of Things and precision positioning technology,huge and high-quality taxi passenger data are obtained during the taxi operation.The record also contains the travel laws of urban residents,urban structure and other social issues.In-depth exploration of taxi passenger data through various data analysis and mining methods is of great significance for intelligent transportation,urban planning,taxi dispatch,and residents’ travel.With the emergence of taxi-driving software such as Didi Mobility and new modes of travel,and the upsurge of machine learning methods,the field of taxi data mining is facing major opportunities and challenges.These passenger data can be mined and modeled through machine learning,and potentially available information can be obtained.Some decisions and predictions can be made using the model.In this paper,New York City taxi passenger data mining combined with machine learning to do the following work:(1)Urban travel hotspots research,based on the point of OD data traffic area divided into hot extraction method to extract the city travel hot spots,and further on the basis of these hot spots build hotspots directed weighted network interactive network,the interaction relationship between the research hotspots,finally,the interactive network community found comparison results by two methods,find out the interaction between the close hot spots.(2)Prediction model construction,based on the OD data extracted from hot spot research combined with the number of passengers,boarding time and taxi fare data for data analysis and feature analysis,build a prediction model and continuously improve the final prediction of fare and arrival time is better Prediction effect.(3)The taxi travel prediction system is implemented,based on urban hotspot area research and fare prediction model construction system,including functions such as travel volume prediction,urban travel hotspot prediction,fare and arrival time prediction,etc.,providing taxi travel and taxi operation reference.This paper uses conventional data mining and analysis methods to conduct research on urban travel hotspots,and then uses machine learning to predict fares and arrival times and apply the research results.The research results of this paper can provide references for urban residents’travel laws,intelligent transportation,urban planning,taxi dispatch,and residents’ travel by taxi.