Analyze And Predict The Hot Spot Regions

Taxi is one of the most popular transports in the world.With the communication technology developing so fast,almost taxis configure GPS system that the drivers will feel convenient and safe.Also the taxis which have configured GPS system will produce huge GPS data every day that many scholars try to analyze the moving data and find de rule that how people moving every day A lot of them make some difference and the success help government to solve the traffic problems.I try to analyze the taxis GPS moving data,and try to find the rule that I can analyze and predict the Hot Spot Regions which people usually wait for the taxis.I want to find and predict the Hot Spot Regions in Beijing and develop the Passengers Recommend System for taxi drivers especially for the novice drivers to carry passengers.Clustering analysis is very useful and popular for scholars to analyze Big Data.There are some popular methods:K-means arithmetic which is based on distance dividing,Single-Linkage arithmetic which is based on gradation level and DBSCAN arithmetic which is based on density.I will do some experiments to compare the three kinds of clustering arithmetic,and then I will try to pick up the best one and then I will apply the arithmetic to my Passengers Recommend System.Then I choose the DBSCAN arithmetic for my system.I will calculate the value how much the area worth to be recommended to the drivers.And I will calculate the distance from the place where the driver want to start to the place where the system recommend,and I will calculate the time the driver need to drive.And I will try to predict the value when the drivers arrive at the area where the system recommends to them.Also I hope the system can help taxi driver to carry more passengers and I also hope the system can less the time that the passengers will wait for the taxis.