EUVIMEDEUVIMED entdecken
Crossref

Construction and Validation of a Neural Network-Based Blended Teaching Quality Evaluation Model

Dongxue Tu

Journal of Big Data and Computing · 2026-06 · 4 (2) · 72-79

DOI: 10.62517/jbdc.202601211 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Big Data and Computing, 2026-06. Verfasst von Dongxue Tu. Die Arbeit nennt 12 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Addressing the persistent challenges in blended teaching quality assessment—overly intricate indicator systems, pronounced subjectivity, and the inadequacy of conventional evaluation approaches—this study takes mathematics courses as a case study and proposes a neural network-based evaluation framework. From offline questionnaires and online teaching platforms, 118 valid samples were obtained. Following min-max normalization, two distinct models were constructed: a standard BP network configured with a 20-7-1 architecture, and a genetic algorithm-enhanced variant. The latter utilized the evolutionary algorithm to optimize initial connection weights and thresholds, resulting in a 20-41-1 topology. MATLAB was employed for both training and testing. The experimental findings show that the standard BP model produced an average error of 4.61 and a relative error of 0.05. By contrast, the model refined through genetic optimization reduced these figures to 1.94 and 0.02, respectively, while also demonstrating a markedly narrower error spread. Its predictive robustness surpassed not only that of the basic BP model but also that of the benchmark algorithms (GA and BSA). The outcomes confirm that the hybrid GA-BP approach achieves significantly higher accuracy and reliability for blended teaching quality evaluation, offering a dependable scientific basis for instructional management decisions.

Zitieren

APA

Tu, D. (2026). Construction and Validation of a Neural Network-Based Blended Teaching Quality Evaluation Model. Journal of Big Data and Computing, 4(2), 72-79. https://doi.org/10.62517/jbdc.202601211

Vancouver

Tu D. Construction and Validation of a Neural Network-Based Blended Teaching Quality Evaluation Model. Journal of Big Data and Computing. 2026;4(2):72-79. doi:10.62517/jbdc.202601211

BibTeX
@article{Tu2026,
  author = {Tu, Dongxue},
  title = {Construction and Validation of a Neural Network-Based Blended Teaching Quality Evaluation Model},
  journal = {Journal of Big Data and Computing},
  year = {2026},
  volume = {4},
  number = {2},
  pages = {72--79},
  doi = {10.62517/jbdc.202601211}
}
Kostenlos starten →