Research map / Probabilistic and scientific modeling / Variational and generative sampling
Distributional Learning of Variational AutoEncoder: Application to Synthetic Data Generation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Increase β from 0.5 to 5 (ablation) | DistVAE · Real tabular datasets (average) | DCR R&S: 0.444 ± 0.250 -> 0.517 ± 0.272 | Helped |
| Increase β from 0.5 to 5 (ablation) | DistVAE · Real tabular datasets (average) | CorrDist: 1.473 ± 0.398 -> 3.113 ± 1.119 | Hurt |
| Increase β from 0.5 to 5 (ablation) | DistVAE · Real tabular datasets (average) | F1: 0.769 ± 0.128 -> 0.679 ± 0.190 | Hurt |
| Increase β from 0.5 to 5 (ablation) | DistVAE · Real tabular datasets (average), k=100 | Attribute disclosure F1: 0.310 ± 0.072 -> 0.232 ± 0.101 | Helped |
| Use an infinite-mixture asymmetric-Laplace decoder with CRPS reconstruction instead of Gaussian decoding | DistVAE versus VAE synthesizer baselines · Real tabular datasets (average), continuous variables | K-S: 0.090 ± 0.065 versus CTGAN 0.133 ± 0.106 | Helped |
| Use an infinite-mixture asymmetric-Laplace decoder with CRPS reconstruction instead of Gaussian decoding | DistVAE versus VAE synthesizer baselines · Real tabular datasets (average) | CorrDist: 1.473 ± 0.398 versus CTGAN 2.180 ± 0.467 | Helped |
| Use DistVAE instead of TVAE | Synthetic tabular data generation · Real tabular datasets (average) | Membership inference accuracy: 0.495 ± 0.019 -> 0.500 ± 0.003 | No clear effect |
Closest papers by meaning
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.90)
- Variational Autoencoder with Implicit Optimal Priors (AAAI 2019 · similarity 0.89)
- Statistical Guarantees for Variational Autoencoders using PAC-Bayesian Theory (NeurIPS 2023 · similarity 0.89)
- Shape your Space: A Gaussian Mixture Regularization Approach to Deterministic Autoencoders (NeurIPS 2021 · similarity 0.88)
- Adversarial Symmetric Variational Autoencoder (NeurIPS 2017 · similarity 0.88)
- Generalization in VAE and Diffusion Models: A Unified Information-Theoretic Analysis (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.