Extreme Learning Machine and Backpropagation Algorithms are used in this study to find out which algorithm is most suitable for knowing student academic achievement. The data about students are explored to get a pattern so that the characteristics of new students can be known every year. The evaluation process of this study uses confusion matrix for the introduction of correctly recognized data and unknown data. Comparison of this algorithm uses student data at the beginning of the lecture as early detection of students who have problems with academics to be anticipated. The variables used are the value of the entrance examination for new students, the first grade IP value, Gender, and Working Status, while the output variable is the quality value as a classification of academic performance. The results of this study state that the Extreme Learning Machine algorithm has a 14.84% error rate lower than Backpropagation 28.20%. From the model testing stage, the most accurate result is the Extreme Learning Machine algorithm because it has the highest accuracy and the lowest error rate.
Copyrights © 2019