GAIN: Missing Data Imputation using Generative Adversarial Nets

Abstract

We propose a novel method for imputing missing data by adapting the well-known Generative Adversarial Nets (GAN) framework. Accordingly, we call our method Generative Adversarial Imputation Nets (GAIN). The generator (G) observes some components of a real data vector, imputes the missing components conditioned on what is actually observed, and outputs a completed vector. The discriminator (D) then takes a completed vector and attempts to determine which components were actually observed and which were imputed. To ensure that D forces G to learn the desired distribution, we provide D with some additional information in the form of a hint vector. The hint reveals to D partial information about the missingness of the original sample, which is used by D to focus its attention on the imputation quality of particular components. This hint ensures that G does in fact learn to generate according to the true data distribution. We tested our method on various datasets and found that GAIN significantly outperforms state-of-the-art imputation methods.

10 pages, 3 figures, 2018 International Conference of Machine Learning

Cited by in corpus (67)

• Synthesizing Tabular Data using Generative Adversarial Networks
• MAIN: Multihead-Attention Imputation Networks
• Missing Data Imputation with Adversarially-trained Graph Convolutional Networks
• Blockchain-based Digital Twins: Research Trends, Issues, and Future Challenges
• Graph Convolutional Networks for Graphs Containing Missing Features
• Maximizing information from chemical engineering data sets: Applications to machine learning
• Processing of missing data by neural networks
• Variational Autoencoder with Arbitrary Conditioning
• On the consistency of supervised learning with missing values
• Spatiotemporal Tensor Completion for Improved Urban Traffic Imputation
• Learning Neural Causal Models from Unknown Interventions
• STING: Self-attention based Time-series Imputation Networks using GAN
• Recovery of Future Data via Convolution Nuclear Norm Minimization
• PermuteAttack: Counterfactual Explanation of Machine Learning Credit Scorecards
• GP-VAE: Deep Probabilistic Time Series Imputation
• Networked Time Series Imputation via Position-aware Graph Enhanced Variational Autoencoders
• Reconstruction of Sentinel-2 Time Series Using Robust Gaussian Mixture Models -- Application to the Detection of Anomalous Crop Development in wheat and rapeseed crops
• Data Imputation with Iterative Graph Reconstruction
• Modeling Network-level Traffic Flow Transitions on Sparse Data
• What's a good imputation to predict with missing values?
• not-MIWAE: Deep Generative Modelling with Missing not at Random Data
• Improving Missing Data Imputation with Deep Generative Models
• Causal Discovery from Incomplete Data: A Deep Learning Approach
• Comparison of Missing Data Imputation Methods using the Framingham Heart study dataset
• Deep Generative Imputation Model for Missing Not At Random Data
• RDIS: Random Drop Imputation with Self-Training for Incomplete Time Series Data
• NAOMI: Non-Autoregressive Multiresolution Sequence Imputation
• Why Not to Use Zero Imputation? Correcting Sparsity Bias in Training Neural Networks
• Instructions and Guide for Diagnostic Questions: The NeurIPS 2020 Education Challenge
• Incomplete Graph Representation and Learning via Partial Graph Neural Networks
• Synthetic Observational Health Data with GANs: from slow adoption to a boom in medical research and ultimately digital twins?
• Linear predictor on linearly-generated data with missing values: non consistency and solutions
• A Comparative Study of Imputation Methods for Multivariate Ordinal Data
• Categorical EHR Imputation with Generative Adversarial Nets
• Are deep learning models superior for missing data imputation in large surveys? Evidence from an empirical comparison
• On the Unreasonable Effectiveness of Feature propagation in Learning on Graphs with Missing Node Features
• Conservative Policy Construction Using Variational Autoencoders for Logged Data with Missing Values
• Astroinformatics based search for globular clusters in the Fornax Deep Survey
• Flow Models for Arbitrary Conditional Likelihoods
• Evaluation of Point Pattern Features for Anomaly Detection of Defect within Random Finite Set Framework
• Unsupervised Data Imputation via Variational Inference of Deep Subspaces
• Robust Multi-Output Learning with Highly Incomplete Data via Restricted Boltzmann Machines
• In-Database Data Imputation
• Causal Discovery from Incomplete Data using An Encoder and Reinforcement Learning
• Generative Adversarial Classification Network with Application to Network Traffic Classification
• Training Generative Adversarial Networks from Incomplete Observations using Factorised Discriminators
• MisConv: Convolutional Neural Networks for Missing Data
• Gradient Importance Learning for Incomplete Observations
• Learning Disentangled Representations of Video with Missing Data
• Blur, Noise, and Compression Robust Generative Adversarial Networks
• Missing Data Imputation for Galaxy Redshift Estimation
• Multiple Organ Failure Prediction with Classifier-Guided Generative Adversarial Imputation Networks
• Variational Auto-Decoder: A Method for Neural Generative Modeling from Incomplete Data
• Causal Inference in medicine and in health policy, a summary
• Networked Time Series Prediction with Incomplete Data via Generative Adversarial Network
• Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows
• Naive imputation implicitly regularizes high-dimensional linear models
• Tomographic Auto-Encoder: Unsupervised Bayesian Recovery of Corrupted Data
• Deep Multi-path Network Integrating Incomplete Biomarker and Chest CT Data for Evaluating Lung Cancer Risk
• Medical data wrangling with sequential variational autoencoders
• EMFlow: Data Imputation in Latent Space via EM and Deep Flow Models
• Missingness Augmentation: A General Approach for Improving Generative Imputation Models
• Using Shapley Values and Variational Autoencoders to Explain Predictive Models with Dependent Mixed Features
• Variational Gibbs Inference for Statistical Model Estimation from Incomplete Data
• Lung Cancer Risk Estimation with Incomplete Data: A Joint Missing Imputation Perspective
• Multiple Imputation for Biomedical Data using Monte Carlo Dropout Autoencoders
• Domain Adaptation with Incomplete Target Domains