Research map / Probabilistic and scientific modeling / Variational and generative sampling
Identifiable Generative models for Missing Not at Random Data Imputation
Chao Ma, Cheng Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Use GINA's identifiable VAE and explicit missingness model instead of Not-MIWAE | VAE-based MNAR imputation · Yahoo! R3 random test set | Test MSE: 1.078 ± 0.000 -> 1.052 ± 0.002 | Improved |
| Model the MNAR mechanism with GINA instead of using the MAR-oriented PVAE | VAE-based imputation · Yahoo! R3 random test set | Test MSE: 1.259 ± 0.003 -> 1.052 ± 0.002 | Improved |
| Use GINA instead of Not-MIWAE for MNAR imputation | VAE-based MNAR imputation · Eedi task 3 test set | Debiased MSE: 0.192 ± 0.000 -> 0.188 ± 0.001 | Improved |
| Use GINA instead of PVAE for MNAR imputation | VAE-based imputation · Eedi task 3 test set | Debiased MSE: 0.194 ± 0.001 -> 0.188 ± 0.001 | Improved |
| Use GINA for active question selection rather than Not-MIWAE | Personalized education active question selection · Eedi task 3 test set | p-value: 0.561 -> 1.01 × 10^-7 | Improved |
| Use GINA for active question selection rather than PVAE | Personalized education active question selection · Eedi task 3 test set | p-value: 0.514 -> 1.01 × 10^-7 | Improved |
Closest papers by meaning
- not-MIWAE: Deep Generative Modelling with Missing not at Random Data (ICLR 2021 · similarity 0.88)
- Generative Modeling under Non-Monotone MAR Missingness via Approximate Wasserstein Gradient Flows (NeurIPS 2026 · similarity 0.86)
- Identification and Estimation for Nonignorable Missing Data: A Data Fusion Approach (ICML 2024 · similarity 0.86)
- Random features models: a way to study the success of naive imputation (ICML 2024 · similarity 0.85)
- On the Identifiability of Hybrid Deep Generative Models: Meta-Learning as a Solution (NeurIPS 2024 · similarity 0.85)
- DiffPuter: Empowering Diffusion Models for Missing Data Imputation (ICLR 2025 · similarity 0.84)
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.
