# Synthetic Data in Education: Empirical Insights from Traditional Resampling and Deep Generative Models

arXiv: https://arxiv.org/abs/2604.21031 · HTML: https://mlsanity.com/paper/2604.21031/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 23 references that are in the map.
- First posted on arXiv 22 Apr 2026.

## Builds on (most cited first, 8 of 8)

- [Modeling Tabular data using Conditional GAN](https://mlsanity.com/paper/1907.00503/) (2019)
- [Generating Multi-label Discrete Patient Records using Generative Adversarial Networks](https://mlsanity.com/paper/1703.06490/) (2017)
- [Efficacy of Synthetic Data as a Benchmark](https://mlsanity.com/paper/2409.11968/) (2024)
- [Sequential Models in the Synthetic Data Vault](https://mlsanity.com/paper/2207.14406/) (2022)
- [Does Differentially Private Synthetic Data Lead to Synthetic Discoveries?](https://mlsanity.com/paper/2403.13612/) (2024)
- [Comparing Synthetic Tabular Data Generation Between a Probabilistic Model and a Deep Learning Model for Education Use Cases](https://mlsanity.com/paper/2210.08528/) (2022)
- [Exploring the Potential of Synthetic Data to Replace Real Data](https://mlsanity.com/paper/2408.14559/) (2024)
- [Creating Artificial Students that Never Existed: Leveraging Large Language Models and CTGANs for Synthetic Data Generation](https://mlsanity.com/paper/2501.01793/) (2025)

## Abstract (arXiv)

Synthetic data generation offers promise for addressing data scarcity and privacy concerns in educational technology, yet practitioners lack empirical guidance for selecting between traditional resampling techniques and modern deep learning approaches. This study presents the first systematic benchmark comparing these paradigms using a 10,000-record student performance dataset. We evaluate three resampling methods (SMOTE, Bootstrap, Random Oversampling) against three deep learning models (Autoencoder, Variational Autoencoder, Copula-GAN) across multiple dimensions: distributional fidelity (Kolmogorov-Smirnov distance, Jensen-Shannon divergence), machine learning utility such as Train-on-Synthetic-Test-on-Real scores (TSTR), and privacy preservation (Distance to Closest Record). Our findings reveal a fundamental trade-off: resampling methods achieve near-perfect utility (TSTR: 0.997) but completely fail privacy protection (DCR ~ 0.00), while deep learning models provide strong privacy guarantees (DCR ~ 1.00) at significant utility cost. Variational Autoencoders emerge as the optimal compromise, maintaining 83.3% predictive performance while ensuring complete privacy protection. We also provide actionable recommendations: use traditional resampling for internal development where privacy is controlled, and VAEs for external data sharing where privacy is paramount. This work establishes a foundational benchmark and practical decision framework for synthetic data generation in learning analytics.