Machine Learning Techniques for Identifying Learning Style and Student’s Performance Prediction: An Approach of Felder Silverman Learning Style Model (FSLSM)

This research proposed a methodology for identifying the student's learning style and student’s performance prediction in the online learning environment using Machine Learning (ML) techniques. Identification of the student's learning style and performance prediction in the teaching and learning environment is important in improving both teaching and learning perspectives. The intention of the research was to investigate about applying Machine Learning Techniques for the identification of the Learning style of the students and the prediction of the student’s performance in an online learning environment based on the Felder Silverman Learning Style (FSLSM) identification model. The significance of this experiment is that the proposed methodology considers the combination of access frequency (f) of course materials and total time (T) students spent on each course activity to reduce the limitations that occur due to accessing the course modules randomly without any preference in the online learning environment in learning style identification. A reusable Moodle time-tracking plugin was created for the data collection procedure. Three-course modules that were created in accordance with the FSLSM model's features were used to prepare a real-time dataset. Seven criteria were chosen, and the features were verified using the Pearson Correlation Coefficient approach. Each of these course modules had 150 enrolled students. Machine learning is a widely used technology for the identification of the learning style and analyzing the data for making predictions. Once the data set was prepared, the data set was preprocessed and applied five Supervised Classification Machine learning algorithms as Decision Tree, Logistic Regression, Random Forest, Support Vector Machine and K-Nearest Neighbors algorithm. The models were evaluated using Accuracy, Precision, Recall and F1 values. Of the five algorithms for Learning Style identification, the Decision Tree classifier algorithm performed with the best average accuracy with 93.5% for Input, 86% for Perception, 89.5 for Processing and 94% for Understanding dimension. For the grade prediction process the Decision Tree algorithm performed with a 96% accuracy level. The models were validated using the K-fold Cross-validation and Standard Deviation values. Mean Squared Error, Bias and Variance values were considered the evaluation of the underfitting or overfitting context of the model. For parameter optimization, the Grid Search Methodology was applied to find the best combination of criterion for the model. Finally, an application was developed for Identifying the Learning Style of the Students and performance prediction using the designed Machine learning model. The Consistency of the ML Model based on the Decision Tree classifier algorithm were evaluated using the results generated through the developed application and the results suggested that consistency for taught machine learning algorithms is often between 85% to 95%, which is an acceptable range. For the grade prediction, the consistency of the models ranged nearly 89%. The results generated by the application for identification of the learning style suggested the combination of learning style for particular students sample as Global-Mild, Visual- Strong, Sensing- Moderate and Reflective-Strong. Identification of these combinations of learning styles assists teachers by giving an insight into which components of the learning content should be improved in the course designing process. One of the limitations is that though how much we encourage the students, some of them do not like to engage in the course works in the online learning environment. These behaviors may lead to difficulties in conducting the data collection process in a precise manner. Providing a mechanism to identify and analyze the factors that impact to increase in the attractiveness of students when reading the course materials or presentation will be one of the main future directions of this teaching and learning research paradigm.