Uncertainty-Aware Deep Classifiers Using Generative Models

Abstract

Deep neural networks are often ignorant about what they do not know and overconfident when they make uninformed predictions. Some recent approaches quantify classification uncertainty directly by training the model to output high uncertainty for the data samples close to class boundaries or from the outside of the training distribution. These approaches use an auxiliary data set during training to represent out-of-distribution samples. However, selection or creation of such an auxiliary data set is non-trivial, especially for high dimensional data such as images. In this work we develop a novel neural network model that is able to express both aleatoric and epistemic uncertainty to distinguish decision boundary and out-of-distribution regions of the feature space. To this end, variational autoencoders and generative adversarial networks are incorporated to automatically generate out-of-distribution exemplars for training. Through extensive analysis, we demonstrate that the proposed approach provides better estimates of uncertainty for in- and out-of-distribution samples, and adversarial examples on well-known data sets against state-of-the-art approaches including recent Bayesian approaches for neural networks and anomaly detection methods.

Rapid Trust Calibration through Interpretable and Uncertainty-Aware AI

“…A common approach is to use Bayesian methods, whereby epistemic uncertainty is captured as uncertainty in the model parameters 33 or as uncertainty in function space using, for example, Gaussian processes. 46 Another promising approach is that of evidential learning, 47,48 whereby inputs are mapped to the parameters of a Dirichlet distribution over classes. Smaller parameter values represent less evidence for a class, producing a broader distribution representing greater epistemic uncertainty.…”

Uncertainty-Aware Reliable Text Classification

“…To ensure the model's generalization to the whole data space, the choice of effective is crucial. Although generative models have achieved success in the CV domain [19,34], they do not apply to discrete text data. We adopt two methods that have achieved success in the NLP domain to get effective OOD regularization: (i) Using auxiliary OOD datasets; (ii) Generating off-manifold adversarial examples.…”

“…A critical finding in [17] is that the diversity of the auxiliary dataset is important. Hu et al [19] report that the methods using diverse examples beat the methods that only use close adversarial examples [14,34] in OOD detection in image classification. Our empirical observations also find that randomly generated sentences (we randomly sample words and concatenate them into fake sentences) do not improve the performance.…”

“…It considers different dimensions of uncertainty, such as vacuity (i.e., lack of evidence) or dissonance (i.e., uncertainty due to conflicting evidence). In the CV domain, Sensoy et al [34] propose evidential neural networks (ENNs) to model the uncertainty of class probabilities based on SL explicitly. An ENN uses the predictions as subjective opinions and learns a function that collects evidence to form the opinions by a deterministic neural network from data.…”

“…An ENN uses the predictions as subjective opinions and learns a function that collects evidence to form the opinions by a deterministic neural network from data. Several works [19,34,51] improve ENNs using regularization or generative models to ensure correct uncertainty estimation towards unseen examples in image classification. However, those methods for continuous feature space are not applicable to the discrete text.…”