Abstract:
This study explores the growth of e-learning in Bangladesh, analyzing its socioeconomic,
environmental, and sustainability implications using survey data. The analysis includes
aspects like internet connectivity, device usage, study hours, course quality, efficacy,
engagement, and affordability. This study applied seven machine learning classification
models—Logistic Regression, Random Forest, K-Nearest Neighbors, Support Vector
Machine, Gaussian Naïve Bayes, Decision Tree, and XGBoost. Random Forest notably
outperformed with 78.22% accuracy in predicting the potential of online education to
replace traditional methods. According to the findings, 98.8% of students had regular
internet access, with 50% reporting extremely stable connections. 60.9% of respondents
use smartphones as their primary device, highlighting the need for mobile-friendly learning
systems. Approximately 50% of students spend less than five hours each week on online
learning platforms such as YouTube and Zoom. 68% of students assess course quality
positively, while 33.7% believe online classes are more effective than traditional methods.
Despite these advantages, obstacles such as technological difficulties, a lack of desire, and
feelings of isolation remain. To overcome these difficulties, the research advises increasing
technical assistance, creating interesting material, and boosting teacher training.
Furthermore, advancements in internet infrastructure and device accessibility are crucial to
the sustained expansion of e-learning. The paper continues with recommendations for
future research, emphasizing the importance of longitudinal studies to measure e-learning's
long-term impact, as well as the potential benefits of hybrid learning models that blend
online and traditional teaching approaches. All things considered, e-learning in Bangladesh
has a lot of promise, but for it to grow sustainably, it must overcome current obstacles.
