Hybridisation of RF(Xgb) to improve the tree-based algorithms in learning style prediction

Haziqah Shamsudin, Maziani Sabudin, Umi Kalsom Yusof


This paper presents hybridization of Random Forest (RF) and Extreme Gradient Boosting (Xgb), named RF(Xgb) to improve the tree-based algorithms in learning style prediction. Learning style of specific users in an online learning system is determined based on their interaction and behavior towards the system. The most common online learning theory used in determining the learning style is the Felder-Silverman’s Learning Style Model (FSLSM). Many researchers have proposed machine learning algorithms to establish learning style by using the log file attributes. This helps in determining the learning style automatically. However, current researches still perform poorly, where the range of accuracy is between 58%-89%. Hence, RF(Xgb) is proposed to help in improving the learning style prediction. This hybrid algorithm was further enhanced by optimizing its parameters. From the experiments, RF(Xgb) was proven to be more effective, with accuracy of 96% compared to J48 and LSID-ANN algorithm from previous literature.


Hybrid; Hyperparameter optimization; Learning style; Online learning; Random Forest

Full Text:


DOI: http://doi.org/10.11591/ijai.v8.i4.pp422-428


  • There are currently no refbacks.

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

IAES International Journal of Artificial Intelligence (IJ-AI)
ISSN/e-ISSN 2089-4872/2252-8938 
This journal is published by the Institute of Advanced Engineering and Science (IAES) in collaboration with Intelektual Pustaka Media Utama (IPMU).

View IJAI Stats