ENHANCED PREDICTION OF STUDENT DROPOUTS USING FUZZY INFERENCE SYSTEM AND LOGISTIC REGRESSION

Abstract
Predicting college and school dropouts is a major problem in educational system and has complicated challenge due to data imbalance and multi dimensionality, which can affect the low performance of students. In this paper, we have collected different database from various colleges, among these 500 best real attributes are identified in order to identify the factor that affecting dropout students using neural based classification algorithm and different mining technique are implemented for data processing. We also propose a Dropout Prediction Algorithm (DPA) using fuzzy logic and Logistic Regression based inference system because the weighted average will improve the performance of whole system. We are experimented our proposed work with all other classification systems and documented as the best outcomes. The aggregated data is given to the decision trees for better dropout prediction. The accuracy of overall system 98.6% it shows the proposed work depicts efficient prediction.

Authors
A. Saranya, J. Rajeswari
Adhiparasakthi Engineering College, India

Keywords
Data Mining, Fuzzy Inference System, Logistic Regression, Decision Trees, Student Dropout
Published By :
ICTACT
Published In :
ICTACT Journal on Soft Computing
( Volume: 6 , Issue: 2 )
Date of Publication :
January 2016

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.