IOSR Journal of Computer Engineering (IOSR-JCE) e-ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 16, Issue 6, Ver. VIII (Nov – Dec. 2014), PP 53-60 www.iosrjournals.org www.iosrjournals.org 53 | Page ID3 Derived Fuzzy Rules for Predicting the Students Acedemic Performance Anita Chaware 1 , Dr. U.A. Lanjewar 2 1 (P.G. Departmet of Computer Science, SNDT WU ,Mumbai, India) 2 (MCA, VMV college of Arts and Science, Nagpur, India) Abstract: This paper presents a technique to use ID3 decision rules to produce fuzzy rules to get the optimize prediction of the students academic performance. In this paper, a the student administrative data for a class is used in order to classify the students final year marks in fuzzy logic prediction . This paper is using the machine learning approach to generate the rules so as to overcome the difficulties in a conventional approach like deriving fuzzy rules base from expert experience. This research provides us with: a way to produce meaningful and simple fuzzy rules; a method to fuzzify ID3-derived rules to deal with many inputs variables; and a de-fuzzification system to get the output in human understandable form. The Id3 tree is generated by the WEKA software and is utilized by the Fuzzy Inference System . A Fuzzy inference system was constructed to give the final crisp output. The ID3 was generated on 300 training data to get the better output. The output of our Fuzzy Student Performance Predictor was then tested on 50 test data to check for the accuracy. Keywords: Student administrative data, Students performance, ID3 Decision tree, Fuzzy rules , Fuzzification, linguistic variables, Membership function , Defuzzification I. Introduction Many statistical, data mining and machine learning approaches have been developed for prediction of students Acedemic Performance by researchers from all over the world [l], [2]. Prediction of Student performance is an active area because it depends on various factors affecting the students performance like the demographic, personal information. Therefore, every researcher has different applied different method to get the optimized results with nearing to 100% accuracy. The higher education has to be improved if the research and skill among the students has to be developed in universities. Many education researchers and instructors have made extensive efforts in constructing effective models to predict student academic performance in a class [1]. This is to reduce the numbers of failures and dropouts of the school. Data mining is one of the major area which help us to predictions and classification. The classification, clustering , etc methods of data mining are capable of classifying the data in data warehouse having large amount of data. They are used to predict categorical class labels and classifies data based on training set, and hence can be used for classifying newly available data[3][4]. Thus it can be outlined as an inevitable part of data mining and has gained more popularity. Instructional intervention is needed as students come from diverse background and to make the learning outcome better. To make instructional intervention, knowledge is acquired by modelling the previous database to find which students need it. One among many existing model is the predictive model to predict the results using factors affecting the results. Predictive models can be useful to the instructor to predict student academic performance and then take some proactive measures like designing an innovative and effective teaching and learning plan to help these academically at-risk students. Additionally, predictive models can reduce the dropout rate of students from relevant courses or programs. Our objective was to explore the students’ academic data of fresher students admitted in First year along with their First yr. marks with a view to classifying their performance using fuzzy logic technique. The paper in its next section 2 i.e. two discusses the scope and method used for this research. Section 3, we discussed briefly about ID3 decision tree and its generation by WEKA software. Section 4 explains how this decision tree can be converted into fuzzy rule. Section 5 we demonstrate the use of fuzzy logic concept to the student database and its how that fuzzy model of students performance predictor is used to predict the student finlal year marks . A 20 student database with its predicted values and actual values with their accuracy results are given in result section 6 Finally concluding the whole work and showing the future scope in section 7. II. Scope And Method Used This study is focused on developing and validating fuzzy mathematical model to predict student academic performance in the university. The predicted results are then validated with the actual results of students to find the accuracy of the model. The Acedemic performance of a student’s is affected by number of factors like previous knowledge, interest, family background, motivation etc.[4]. In this paper we have considered the factors like the student’s first year result and their class attendance, their HSC score Graduation