Predicting Learning Result of Learner in E-learning Course with Feature Selection Using SVM

Kitanaka, Yuki; Takeuchi, Kazuhiro; Hirokawa, Sachio

doi:10.1145/3175536.3175567

Cited by 5 publications

(1 citation statement)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Untuk meningkatkan hasil sistem pembelajaran berbasis e-learning disuatu perguruan tinggi, tentu perlu dilakukan evaluasi dengan melihat aktivitas log pengampu mata kuliah serta aktivitas log user ataupun mahasiswa, dalam hal ini kami melakukan penelitian dengan memprediksi hasil dari suatu pembelajaran di perguruan tinggi. Data yang kami dapatkan yang mengenai log aktivitas didapat dari lembaga pendidikan, data semacam ini yang disebut dengan Educational Data Mining [1]. Sistem pembelajaran e-learning terus berkembang pesat, didukung oleh peningkatan kecanggihan teknologi informasi dan dengan pemahaman yang lebih baik tentang bagaimana membuat konten yang lebih efektif.…”

Section: Pendahuluanunclassified

Prediction of Successful Elearning Based on Activity Logs with Selection of Support Vector Machine based on Particle Swarm Optimization

Saputra

Putri²,

Indriyanti³

2019

IJAIDM

View full text Add to dashboard Cite

Prediction is a systematic estimate that identifies past and future information, we predict student learning success with e-learning based on a log of student activities. In this study, we use the Support vector Machine (SVM) method, which is compared with Particle Swarm Optimization. The problem with this algorithm is that the SVM has a very good generalization that can solve a problem. However, some of the attributes in the data can reduce accuracy and add complexity to the SVM algorithm. For this reason, attribute selection for existing data is needed, therefore Particle Swarm Optimization (PSO) method is applied for the right attribute selection in determining the success of elearning learning based on student activity logs, because the PSO method can improve accuracy in determining selection of attributes. The SVM algorithm produces an accuracy value of 88.00% and AUC with a value of 0.8120, while with SVM Based on PSO the accuracy value is 88.50% and the AUC value is 0.8460. Therefore, there is an increase from the result of an accuracy value of 0.50% and an AUC value of 3.40%, and then the result is in good classification.

show abstract

Section: Pendahuluanunclassified