Research And Application Of Campus WiFi User Trajectory Data Clustering Algorithm

With the coverage of school wireless network and the widespread popularization of mobile intelligent terminals,teachers and students generate a large amount of wireless network data in the process of using wireless network,which contains the campus trajectory data of teachers and students.In order to efficiently extract valuable trajectory data from these complex data,mine trajectory movement rules,conduct cluster analysis and apply them to actual education and teaching,it is particularly important to choose an appropriate trajectory clustering algorithm.The main work is as follows:(1)First,using the distance formula combining Hausdorff distance and Frechet distance,The optimization is applied to the neighborhood distance calculation of the DBSCAN algorithm;secondly,the grid method is introduced to search the neighborhood range to reduce the time complexity of the algorithm;finally,an improved dynamic feature trajectory extraction method is proposed to adapt to the complex trajectories of the campus,and extract the cluster flexibly and accurately.The feature trajectory after the class can effectively and quantitatively describe the overall motion law of the trajectory.Taking campus Wi Fi data as an example,the experimental results show that compared with other clustering algorithms,the algorithm proposed in this paper is more suitable for trajectory classification in terms of operating efficiency,storage space,accuracy,cohesion and separation,and is superior to other algorithms.(2)Carrying out application research on HF-DBSCAN algorithm.Fin BI is used to display the trajectory clustering path and feature trajectory extraction effect clearly and intuitively.Using Superset to connect to the database,query the trajectory clustering data according to the actual application scenario,visualize the query results,and further explore the rules of student group behavior and individual behavior.Using the improved algorithm to train the trajectory prediction model to improve the prediction accuracy.The results show that the improved clustering algorithm proposed in this paper has better application effect.The research in this paper is helpful to excavate the movement rules of trajectories on campus,enrich and detail the research and application of trajectory clustering,and provide ideas for related scientific research work.