EUVIMEDEUVIMED entdecken
Crossref

A digital model for predicting the academic success of international students

Evgeny А. Kukuev, Vladimir A. Mishhenko, Egor K. Karpov

Perspectives of science and Education · 01.04.2026 · 80 (2) · 572-588

DOI: 10.32744/pse.2026.2.35 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Perspectives of science and Education, online erschienen am 01.04.2026. Verfasst von Evgeny А. Kukuev und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Increasing the heterogeneity of the student environment, in particular due to an increase in the number of international students, requires an organized, operational analysis of educational data to support them. The possibility of anticipating the risks of educational situations (low academic performance, expulsion, etc.) ensures the safety of the contingent and the building of an individual educational trajectory. The purpose of the work: to analyze the effectiveness of a digital model for processing digital educational data (attendance, academic performance) to predict the success of the educational process. Methodology and methods. The study was based on aggregated data on 14,820 students of Tyumen State University studying in digital educational disciplines. The following methods were used in the study: Pandas; Sklearn (scikit-learn); implementation of clustering (KMeans, AgglomerativeClustering), regression (RandomForestRegressor, MultiOutputRegressor), data standardization (StandardScaler), model evaluation (r2_score, silhouette_score) and statistical checks (chi2_contingency); UMAP (Uniform Manifold Approximation and Projection); KMeans; HDBSCAN; AgglomerativeClustering; Interpolation; Regression (RandomForestRegressor). KEYWORDS Results. Statistically significant is the higher academic performance of Russian students in most disciplines compared to foreign ones. The hypothesis of higher attendance among Russian students has been confirmed. Based on the analysis of average curves and cluster centers, the types of student behavior were interpreted: Stable; Disappearing; Skipping; Jerky.

Conclusion

The results obtained indicate the relevance of using digital tools in the analysis of educational situations. The revealed trends towards lower attendance by international students may be an indicator of potential risks of learning success. The creation of digital products that facilitate the rapid analysis of educational data and the forecasting of the success of the educational process on this basis provides an opportunity to anticipate negative situations in the education of students, including from other countries. Using objective data increases the effectiveness of the educational process, and hence the success of students' professional development.

Zitieren

APA

Kukuev, E. А., Mishhenko, V. A., & Karpov, E. K. (2026). A digital model for predicting the academic success of international students. Perspectives of science and Education, 80(2), 572-588. https://doi.org/10.32744/pse.2026.2.35

Vancouver

Kukuev EА, Mishhenko VA, Karpov EK. A digital model for predicting the academic success of international students. Perspectives of science and Education. 2026;80(2):572-588. doi:10.32744/pse.2026.2.35

BibTeX
@article{Kukuev2026,
  author = {Kukuev, Evgeny А. and Mishhenko, Vladimir A. and Karpov, Egor K.},
  title = {A digital model for predicting the academic success of international students},
  journal = {Perspectives of science and Education},
  year = {2026},
  volume = {80},
  number = {2},
  pages = {572--588},
  doi = {10.32744/pse.2026.2.35}
}
Kostenlos starten →