A Study Of College Curricula-variable Based On Machine Learning

With the introduction of the Internet plus,it is not only in the field of expertise,but with the gradual integration of traditional industries.Internet services and applications under the new economy has penetrated into all aspects of people's lives,the original activity of human being is becoming more convenient,faster and more intelligent.Especially cloud computing,big data technology applications,for the user to find the information they need in the massive data,to some extent,the user is not only to use the information,also the maker of information,the communicator,at the same time,information push has become a hot topic in the industry and academia,personalized information push in the E-commerce Platform,Digital Library Technology,Internet Online Learning has become one of the most developed technologies.Push service which has achieved initial success in various industries,but in the field of education is relatively slowly.The reasons are manifold,educational information are complex,involving a lot of privacy issues,digital campus is not mature enough and so on.In this thesis,the online courses selection in college has so many types and numbers,study on the blindness of students' course selection,aimed at the "12th five-year plan" put forward "wisdom campus",design of push algorithm based on Hybrid recommendation in machine learning,trying to provide a reference for students to choose the right course,this paper puts forward an idea of intelligent education.The main contents of this paper are as follows five:I.First,this thesis defines several important concepts of college course selection system,make sure the research of this thesis is " curricula-variable ",and put forward many problems of elective courses.The next,combing the history and current situation of information push,and putting forward to the application of information push method to solve the blindness of elective courses.II.This thesis makes a comprehensive comparison of the information filtering technology in information push,and also makes a deep analysis on the hybrid push algorithm based on machine learning,collaborative filtering algorithm and content-based recommendation algorithm.Clear the limitations of various technologies and combined with the study of this paper,the design method of curriculum push algorithm to reduce blindness of students' course selection.IV.The content based recommendation algorithm is used as the main idea of curriculum classification,and the classification model is constructed.The specific method is to use text mining process “talent training program of electronic commerce” overview.V.In this thesis,we propose a collaborative filtering recommendation algorithm,which is based on the user's recommendation technology as the main idea of student clustering.The students will be quantified and converted into the data format,the process of clustering similarity.VI.Design algorithm.According to the complexity of the educational data and the actual situation of the e-commerce of S University,the algorithm is designed,and the mixed recommendation is used as the whole structure,and the collaborative filtering is the main content,based on content.Finally,experimental data are used to verify the accuracy of the algorithm,and the results are analyzed.