Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
Explore this paper's citation graph
Summary
This work proposes an alternative to Bayesian NNs that is simple to implement, readily parallelizable, requires very little hyperparameter tuning, and yields high quality predictive uncertainty estimates.
- Type
- preprint
- Published
- 2016-12-05
- Cited by
- 8,412
- References
- 63
- Access
- Open access
- OpenAlex
- https://openalex.org/W2560321925
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:6294674
Keywords
Hyperparameter, Computer science, Robustness (evolution), Machine learning, Bayesian probability
References
- Stacked generalization
- A Bayesian encourages dropout
- Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning
- Strictly Proper Scoring Rules
- Multiple Classifier Systems
- Estimating the mean and variance of the target probability distribution
- Stochastic Expectation Propagation
- Bayesian dark knowledge
- The Comparison and Evaluation of Forecasters.
- Efficient Estimation of Word Representations in Vector Space
- Rectified Linear Units Improve Restricted Boltzmann Machines
- Intriguing properties of neural networks
- Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks
- Distilling the Knowledge in a Neural Network
- The Well-Calibrated Bayesian
- Strictly Proper Scoring Rules, Prediction, and Estimation
- Extremely randomized trees
- VERIFICATION OF FORECASTS EXPRESSED IN TERMS OF PROBABILITY
- Dropout: a simple way to prevent neural networks from overfitting
- Healing the relevance vector machine through augmentation
Cited by
- Uncertainty-aware performance assessment of optical imaging modalities with invertible neural networks
- Mixture density networks for distribution and uncertainty estimation
- Multiplicative Normalizing Flows for Variational Bayesian Neural Networks
- UCB EXPLORATION VIA Q-ENSEMBLES
- On Calibration of Modern Neural Networks
- Safe Visual Navigation via Deep Learning and Novelty Detection
- Uncertainty-Aware Learning from Demonstration Using Mixture Density Networks with Sampling-Free Variance Modeling
- Distance-based Confidence Score for Neural Network Classifiers
- Convolutional neural networks that teach microscopes how to image
- Uncertainty averse pushing with model predictive path integral control
- Overpruning in Variational Bayesian Neural Networks
- Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification
- Uncertainty Estimation via Stochastic Batch Normalization
- Learning uncertainty in regression tasks by artificial neural networks
- A Scalable Laplace Approximation for Neural Networks
- Gradient conjugate priors and deep neural networks
- High-Quality Prediction Intervals for Deep Learning: A Distribution-Free, Ensembled Approach
- Diversity regularization in deep ensembles
- Uncertainty Estimates for Optical Flow with Multi-Hypotheses Networks
- Active Model Learning and Diverse Action Sampling for Task and Motion Planning
Related papers
- Deep Residual Learning for Image Recognition
- On calibration of modern neural networks
- Weight Uncertainty in Neural Networks
- Learning Multiple Layers of Features from Tiny Images
- Practical Variational Inference for Neural Networks
- Weight Uncertainty in Neural Network
- Gradient-based learning applied to document recognition