Analysis of Student Performance using Data Mining

Abstract

The aim of this thesis paper is to analyze student performance using data mining. Data mining is the process of prediction, extracting data. Prediction regarding student performance can help a student to take decision. It can help not only the current students but also the future students, to take decision. In this way they can avoid poor performance which will help to enhance their performance. This is also a guideline to take decision. To understand student performance, a survey was conducted by Military InstituteofScience Technology(MIST)withthesupportfromtheCSEdepartmentand the peer learners of different classes. The data collected from survey was normalized, validatedandrevalidated. Afterthoroughinvestigationonthesurveydata,basedonstatisticalanalysistechniques,differntobservationswererecordedintheformofgraphical illlustrationinordertofindtherelations. Theexperimentalanalysisofthedatathrough result form the survey was satisfactory whicn led towards further study. In order to proceedfurtherthroughdataminingbasedontheunderstandingofthesurvey,datawas collected form the central databse of MIST where the main aim was to relate CGPA and student performance. We investigated different properties of the data; collected and developed a classification hypothesis in order to apply data mining algorithms. In this reasearch a machine learning tool called WEKA develop by the university of New Zealand was used for testing different algorithms on the data. It is to be noted successful appliaction of data minig algorithms in weka requires careful analysis of the source data and develops very specific classes using a very specialized format called ARFF (Attribute related file format) format, which defines attributes a targeted class based on data observations. The experimental results are validated against test data and intersting co-relations are observed. In the future further regorous study to match between demographic data and academic data will lead to much determining factors in order to predict the student performance.