The Prediction Of Students’ Location Based On The Data Of The Campus Smart Card

The development of contemporary science and technology is gradually moving towards the path of informatization.Research on the development of big data is advancing the transformation of digital campus to smart campus.Among which,the application of student campus card and the deployment of campus card system make the work and life of teachers and students in colleges and universities more convenient.Campus cards bring convenience and efficiency to the daily life,and generates a large amount of user’s behavior log data.College administrators is studying the behavior data generated by students in school life as well as using machine learning and big data theory to analyze the growth track of students,and to provide more information for college teaching,student management,campus life,scientific and intelligent decisionmaking services.Based on the behavior data of college students in school life,this article portrays the behavior portraits of the students and predicts the locations of the students’ behavior activities.Through the analysis of the behavior data of the students’ campus card life in the school,the subdivision for behavior problems of the students has proceeded,whereby the behavior model is constructed.1)Integrate student’s personalized behavior factors,use machine learning method to analyze student’s behavior rule,construct time series features with student’s personality information,so as to establish LSTM network model to realize the prediction of student’s future credit card consumption location.2)Improve the structure of existing models based on deep learning of algorithms.It is combined with the advantages of convolutional neural networks for feature extraction,meanwhile the Bi-directional LSTM network is used to process time series data,so that the above two deep learning models could be integrated to improve the accuracy of problem prediction.3)Compare the advantages and disadvantages between the improved method model and the traditional method model,compare and analyze the experimental results.The experimental comparison indicated that the improved method model proposed in this paper would improve the accuracy rate compared with the original use model.It provides certain reference opinions for universities to more conveniently control student flow and improve campus environment,and has important application significance for better student management in colleges and universities.