Analysis Of Middle School Students' Performance Based On Fuzzy Inference Model

As the main form of talent selection,examinations often occur in our lives.From the student's age to professional life,test scores are important indicators of our ability.Especially in the high school years,a large amount of grade data filled the gap between students and teachers such as graduation results,final grades,quiz scores.All of us know how important is the achievement.However,the use of academic performance by teachers and students is limited to simple descriptive statistics.This is not just a waste of this data resource.It is still a lack of scientific,comprehensive,and targeted timely advice for teacher management students and even student self-directed learning.Therefore,it is very urgent to find a data mining technology that enables a wide range of applications,simple operation,and learning ability to analyze student achievement and provide useful information hidden behind a variety of redundant data.This paper selected a fuzzy reasoning model that is close to human thinking to process data.With statistical R software,calling R packages solves the implementation of fuzzy clustering,fuzzy reasoning,and multiple regression.This paper first carried out fuzzy c-means clustering and k-means clustering algorithms for the results of the three examinations for middle school students(three distances)and then attributed 4 categories to 4 levels:excellent,good,medium,and poor;The characteristics of each category were characterized to obtain the overall distribution and the specific characteristics of students in each category.In combination with the discussion of category differences,each category of students is given specific advantages and short-board disciplines so that teachers can make targeted teaching.The weighted average of the four clustering results was used to calculate the distribution of the total scores and analyze them.It was found that some students had a small crossover between the categories and the distribution of scores,accounting for about 10%.For this group of students as a special one to make warnings,focus on observation.Finally,from the analysis of category changes in the three stages,combined with the distribution of special classes,the relationship between class and categories is discussed comprehensively.From the above perspectives,the analysis of the overall category of the total scores,the analysis of the comprehensive scores of arts and sciences,and the attention to the abnormal data were handled.After dealing with the distribution of results,the model under T(r)fuzzy reasoning and regression analysis were used to predict the performance of the next stage without intervention.After comparing the results of the two forecast results,the weighted average of the two is used as the final forecast result.Comparing the ninth-grade predicted results with the other two-stage results,we can identify possible special populations.These students who are at risk are given targeted instruction and learning suggestions so that they can make timely adjustments.