Today, blended learning is widely carried out in many colleges. Different online learning platforms have accumulated a large number of fine granularity records of students’ learning behavior, which provides us with an excellent opportunity to analyze students’ learning behavior. In this paper, based on the behavior log data in four consecutive years of blended learning in a college’s programming course, we propose a novel multiclassification frame to predict students’ learning outcomes. First, the data obtained from diverse platforms, i.e., MOOC, Cnblogs, Programming Teaching Assistant (PTA) system, and Rain Classroom, are integrated and preprocessed. Second, a novel error-correcting output codes (ECOC) multiclassification framework, based on genetic algorithm (GA) and ternary bitwise calculator, is designed to effectively predict the grade levels of students by optimizing the code-matrix, feature subset, and binary classifiers of ECOC. Experimental results show that the proposed algorithm in this paper significantly outperforms other alternatives in predicting students’ grades. In addition, the performance of the algorithm can be further improved by adding the grades of prerequisite courses.
Surface roughness is a critical indicator for assessing the quality and characteristics of workpieces, the accurate prediction of which can significantly enhance production efficiency and product performance. Data-driven methods are efficient ways for predicting surface roughness in polishing processes, which generally depend on large-scale samples for model training. However, obtaining an adequate amount of training data during the polishing process can be challenging due to constraints related to cost and efficiency. To address this issue, a novel surface roughness prediction model, named VSG-FC, is proposed in this paper that integrates Genetic Algorithm-driven Virtual Sample Generation (GA-VSG) and Genetic Programming-driven Feature Construction (GP-FC) to overcome data scarcity. This approach optimizes the feature space through sample augmentation and feature reconstruction, thereby enhancing model performance. The VSG-FC method proposed in this paper has been validated using data from two polishing experiments. The results demonstrate that the method offers significant advantages in both the quality of the generated virtual samples and prediction accuracy. Additionally, the proposed model is explainable and could successfully identify key influencing machining factors.
In recent years, blended learning models combining online and offline instructional methods have been widely implemented in higher education institutions. Online learning platforms accumulate a large number of fine-grained records of students' learning behavior, offering an excellent opportunity to analyze students' learning behavior. Firstly, a year's worth of student behavior log data was collected from the “Java Language Programming” blended learning course at a college. Secondly, we propose an interpretable grade prediction method based on the error correcting output code (ECOC) with Shapley theory. This approach employs Recursive Feature Elimination (RFE) to select features from the original features, and creates an ECOC classification matrix to predict student performance. Finally, the Shapley values for each feature were used to conduct an explainable analysis of the prediction results. Experimental results show that the proposed method outperforms other classical classification algorithms. This study provides a visual and educational analysis of the major factors affecting academic performance, thereby offering educational administrators more intuitive decision support.