Application Of Matrix Completion Method In University Comprehensive Ranking Based On The Fourth Round Of Subject Assessment And Simple Image Processing

It is an era of large data when precision is greatly required.Problems frequently occur at work and daily lives,like ranking of all students of the same class,who cannot be ranked by add the scores of all subjects because of their inconsistent of course selection even if they share the same credits.Another example is the popularity ranking of 100 movies released in 2017,where it is hoped to determine by collecting questionnaires scored by fans.Unfortunately,many fans failed to watch all of the 100 movies,so it attracts people’s attention how to scientifically determine the ranks of movies on the basis of these incomplete questionnaires.Problems like them involve a wide range of fields,which may be encountered in machine learning,system recommendation,image processing and video system processing.In recent years,mathematicians have invented a brand-new method,matrix completion method,to solve such problems.In this paper,the author will briefly describe the matrix completion method and apply it to deal with the fourth round of discipline assessment data released by the Degree Center of the Ministry of Education on December 28,2017,where the author sets up a variation model,selects the SVT algorithm for matrix-completion of the scores of disciplines in 490 universities,all the scores of disciplines in these universities,obtains the overall ranking of the above-mentioned 490 universities by the order of the total scores and finally compare with the rankings from Alumni Association and Mr.Wu Shulian,and the result shows that is similar.At the same time,this method also comes as an answer the questions that which subject we recommend should be set up if a university intends to expand the number of disciplines so as to achieve the optimal layout of disciplines.In this paper,matrix completion is also applied to simple contaminated gray-scale image processing,where the corresponding matrix is filled to the contaminated image and converts them into an image,which plays the role in image restoration.It can be seen that the matrix completion is a very effective method that can be applied to a number of occasions.