Reliability analysis in curriculum development for social science education driven by machine learning
This research aimed at applying machine learning models to improve reliability in the development of social science courses. Lasso was considered a regression technique that selected relevant features for prediction but eliminated redundant features. The model was subsequently trained with sub-sampl...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-05-01
|
| Series: | Alexandria Engineering Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1110016825002261 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This research aimed at applying machine learning models to improve reliability in the development of social science courses. Lasso was considered a regression technique that selected relevant features for prediction but eliminated redundant features. The model was subsequently trained with sub-sample data containing 80% of the data while this part was further tested with 20% for verifying the generalizability and robustness of the model. Performance evaluation was conducted on the linear regression, random forest and artificial neural networks (ANN) through statistical metrics such as root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). The linear regression proved to produce an accurate result in terms of RMSE and MAE, meaning it can behave extremely promisingly in predicting the course trends. However, the high MAPE values support the same findings, confirming its predictive performance. The highest error records of random forest emerged because the model experienced difficulties processing linear data patterns and yet failed to produce efficient models. This is a strong justification for advocating innovative mechanisms for planners to use machine learning techniques for the design of effective data-based, credible, and efficient social science curriculum through the use of the linear regression and ANN. |
|---|---|
| ISSN: | 1110-0168 |