.png){kind=icon}
.png){kind=icon}
Return
Spark framework for student grade classification using ResNext-Xception model
.png){kind=icon}
PRE
AI
DOI：10.1080/03610918.2026.2644594.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The significance of educating the next generation in the comprehension basics of future technological and scientific innovations will force an extensive economic and social pattern that cannot be overstressed. Moreover, artificial intelligence (AI) is considered as one such advanced technology, particularly machine learning (ML) approaches. Even though learning analytics and educational data mining have experienced a rise in utilization and exploration, they are still tricky to accurately describe. Application of deep learning (DL) techniques for classifying student grades has been well-received by researchers. Therefore, key intention of this work is to suggest a new hybrid DL approach to classify the student grades in Spark framework. Primarily, the data is subjected to data splitting by employing deep embedded clustering with data augmentation (DEC-DA). Moreover, DEC-DA model is trained by Tyrannosaurus optimization algorithm (TOA). Subsequently, it is transferred to the pre-processing phase, which is performed by data cleaning and data scaling. Spearman's rank correlation coefficient is employed for feature selection. Ultimately, the student grade classification is accomplished by ResNext-Xception, which is an integration of ResNext and Xception models. Moreover, experimental analysis is conducted for the proposed model by considering performance metrics, like accuracy, sensitivity and specificity, where presented approach reached utmost values of 0.959, 0.965, and 0.949.
Keywords:
Deep learning
Educational data mining
Optimization algorithm
Spark framework
Student grade classification
Journal
C
IF:
0.8
Papers:
163
Citations:
4.7K