A COMPARISON BETWEEN SIMPLE LINEAR REGRESSION AND RADIAL BASIS FUNCTION NEURAL NETWORK (RBFNN) MODELS FOR PREDICTING STUDENTS’ ACHIEVEMENT

Sunarto, Andang and Yuniarti, Suci (2014) A COMPARISON BETWEEN SIMPLE LINEAR REGRESSION AND RADIAL BASIS FUNCTION NEURAL NETWORK (RBFNN) MODELS FOR PREDICTING STUDENTS’ ACHIEVEMENT. UMS. pp. 299-308.

Preview

Text 3.UMS_UNJ_2014_AComparation Analysis of Linear Regression and Artificial Neural Networks.pdf Download (3MB) | Preview

Abstract

This paper presents an approach for predicting student achievements using statistics and artificial neural networks (ANN), namely linear regression and radial basis function neural network (RBFNN) methods. The data is gained from 108 students from mathematics department in Islamic University, Bengkulu, Indonesia. The results of measurement are then compared to the value of the mean of square error (MSE). The results show that MSE 0.076 with model Y = 3.193 + 0.002 for linear regression and MSE 0.003, model Y = (1)T + (0.0021) with sum-squared error goal 0.01, and spread 1 for the RBFNN. The smallest MSE value indicates a good method for accuracy. Therefore, the RBFNN model illustrates the proposed best model to predict students’ achievement.

Actions (login required)

View Item