Comparison of Data Imputation Performance in Deep Generative Models for Educational Tabular Missing Data

Computer Science

• Deep Generative Model 100%
• Denoising Diffusion Probabilistic Models 100%
• Sampling Technique 60%
• Generative Adversarial Networks 20%
• Performance Model 20%
• Critical Information 20%
• Educational Data Mining 20%
• Extreme Gradient Boosting 20%
• Class Imbalance 20%
• Classification Task 20%
• Learning Analytics 20%
• Variational Autoencoder 20%

Engineering

• Generative Model 100%
• Sampling Technique 100%
• Classification Task 33%
• Variational Autoencoder 33%
• Xgboost 33%

Mathematics

• Data Imputation 100%
• Autoencoder 33%
• Predictive Modeling 33%