Timezone: »
This paper proposes a fast and scalable method for uncertainty quantification of machine learning models' predictions. First, we show the principled way to measure the uncertainty of predictions for a classifier based on Nadaraya-Watson's nonparametric estimate of the conditional label distribution. Importantly, the approach allows to disentangle explicitly \textit{aleatoric} and \textit{epistemic} uncertainties. The resulting method works directly in the feature space. However, one can apply it to any neural network by considering an embedding of the data induced by the network. We demonstrate the strong performance of the method in uncertainty estimation tasks on a variety of real-world image datasets, such as MNIST, SVHN, CIFAR-100 and several versions of ImageNet.
Author Information
Nikita Kotelevskii (Skolkovo Institute of Science and Technology)
Alexander Fishkov (Skolkovo Institute of Science and Technology)
Kirill Fedyanin (Skolkovo Institute of Science and Technology)
Aleksandr Petiushko (Lomonosov Moscow State University)
Maxim Panov (Skoltech)
More from the Same Authors
-
2022 : Distributional deep Q-learning with CVaR regression »
Mastane Achab · REDA ALAMI · YASSER ABDELAZIZ DAHOU DJILALI · Kirill Fedyanin · Eric Moulines · Maxim Panov -
2022 Poster: Smoothed Embeddings for Certified Few-Shot Learning »
Mikhail Pautov · Olesya Kuznetsova · Nurislam Tursynbek · Aleksandr Petiushko · Ivan Oseledets -
2022 Poster: Nonparametric Uncertainty Quantification for Single Deterministic Neural Network »
Nikita Kotelevskii · Aleksandr Artemenkov · Kirill Fedyanin · Fedor Noskov · Alexander Fishkov · Artem Shelmanov · Artem Vazhentsev · Aleksandr Petiushko · Maxim Panov -
2022 Poster: FedPop: A Bayesian Approach for Personalised Federated Learning »
Nikita Kotelevskii · Maxime Vono · Alain Durmus · Eric Moulines