MIWAE: Deep Generative Modelling and Imputation of Incomplete Data

Keyphrases

• Deep Generative Modeling 100%
• Incomplete Data 100%
• Single Imputation 50%
• Importance Weighted Autoencoder 50%
• State-of-the-art Techniques 25%
• Tight Lower Bound 25%
• Log-likelihood Ratio 25%
• Simple Technique 25%
• Training Set 25%
• Computation Overhead 25%
• Missing Data 25%
• Multiple Imputation 25%
• Monte Carlo Method 25%
• Convolutional 25%
• Continuous Data 25%
• Handling Missing Data 25%
• Nonignorable Missingness 25%
• MNIST 25%

Mathematics

• Incomplete Data 100%
• Single Imputation 50%
• Training Set 25%
• Missing At Random 25%
• Handling Missing Data 25%
• Log Likelihood 25%
• Multiple Imputation 25%
• Monte Carlo Technique 25%
• Observed Data 25%
• Continuous Data 25%