Model confidence or uncertainty is critical in autonomous systems as they directly tie to the safety and trustworthiness of the system. The quantification of uncertainty in the output decisions of deep neural networks (DNNs) is a challenging problem. The Bayesian framework enables the estimation of the predictive uncertainty by introducing probability distributions over the (unknown) network weights; however, the propagation of these high-dimensional distributions through multiple layers and non-linear transformations is mathematically intractable. In this work, we propose an extended variational inference (eVI) framework for convolutional neural network (CNN) based on tensor Normal distributions (TNDs) defined over convolutional kernels. Our proposed eVI framework propagates the first two moments (mean and covariance) of these TNDs through all layers of the CNN. We employ first-order Taylor series linearization to approximate the mean and covariances passing through the non-linear activations. The uncertainty in the output decision is given by the propagated covariance of the predictive distribution. Furthermore, we show, through extensive simulations on the MNIST and CIFAR-10 datasets, that the CNN becomes more robust to Gaussian noise and adversarial attacks.
more »
« less
Sinh-arcsinh-normal distributions to add uncertainty to neural network regression tasks: Applications to tropical cyclone intensity forecasts
Abstract A simple method for adding uncertainty to neural network regression tasks in earth science via estimation of a general probability distribution is described. Specifically, we highlight the sinh-arcsinh-normal distributions as particularly well suited for neural network uncertainty estimation. The methodology supports estimation of heteroscedastic, asymmetric uncertainties by a simple modification of the network output and loss function. Method performance is demonstrated by predicting tropical cyclone intensity forecast uncertainty and by comparing two other common methods for neural network uncertainty quantification (i.e., Bayesian neural networks and Monte Carlo dropout). The simple approach described here is intuitive and applicable when no prior exists and one just wishes to parameterize the output and its uncertainty according to some previously defined family of distributions. The authors believe it will become a powerful, go-to method moving forward.
more »
« less
- Award ID(s):
- 2019758
- PAR ID:
- 10422653
- Date Published:
- Journal Name:
- Environmental Data Science
- Volume:
- 2
- ISSN:
- 2634-4602
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Each year a growing number of wind farms are being added to power grids to generate sustainable energy. The power curve of a wind turbine, which exhibits the relationship between generated power and wind speed, plays a major role in assessing the performance of a wind farm. Neural networks have been used for power curve estimation. However, they do not produce a confidence measure for their output, unless computationally prohibitive Bayesian methods are used. In this paper, a probabilistic neural network with Monte Carlo dropout is considered to quantify the model or epistemic uncertainty of the power curve estimation. This approach offers a minimal increase in computational complexity and thus evaluation time. Furthermore, by adding a probabilistic loss function, the noise or aleatoric uncertainty in the data is estimated. The developed network captures both model and noise uncertainty which are found to be useful tools in assessing performance. Also, the developed network is compared with the existing ones across a public domain dataset showing superior performance in terms of prediction accuracy. The results obtained indicate that the developed network provides the quantification of uncertainty while maintaining accurate power estimation.more » « less
-
Lai, Yuan (Ed.)Mistrust is a major barrier to implementing deep learning in healthcare settings. Entrustment could be earned by conveying model certainty, or the probability that a given model output is accurate, but the use of uncertainty estimation for deep learning entrustment is largely unexplored, and there is no consensus regarding optimal methods for quantifying uncertainty. Our purpose is to critically evaluate methods for quantifying uncertainty in deep learning for healthcare applications and propose a conceptual framework for specifying certainty of deep learning predictions. We searched Embase, MEDLINE, and PubMed databases for articles relevant to study objectives, complying with PRISMA guidelines, rated study quality using validated tools, and extracted data according to modified CHARMS criteria. Among 30 included studies, 24 described medical imaging applications. All imaging model architectures used convolutional neural networks or a variation thereof. The predominant method for quantifying uncertainty was Monte Carlo dropout, producing predictions from multiple networks for which different neurons have dropped out and measuring variance across the distribution of resulting predictions. Conformal prediction offered similar strong performance in estimating uncertainty, along with ease of interpretation and application not only to deep learning but also to other machine learning approaches. Among the six articles describing non-imaging applications, model architectures and uncertainty estimation methods were heterogeneous, but predictive performance was generally strong, and uncertainty estimation was effective in comparing modeling methods. Overall, the use of model learning curves to quantify epistemic uncertainty (attributable to model parameters) was sparse. Heterogeneity in reporting methods precluded the performance of a meta-analysis. Uncertainty estimation methods have the potential to identify rare but important misclassifications made by deep learning models and compare modeling methods, which could build patient and clinician trust in deep learning applications in healthcare. Efficient maturation of this field will require standardized guidelines for reporting performance and uncertainty metrics.more » « less
-
more » « less
Abstract Evaluating the exact first derivative of a feedforward neural network (FFNN) output with respect to the input feature is pivotal for performing the sensitivity analysis of the trained neural network with respect to the inputs. In this paper, a novel method is presented that computes the analytical quality first derivative of a trained feedforward neural network output with respect to the input features without the need for backpropagation. To this end, the complex step derivative approximation is illustrated, and its implementation in the framework of the feedforward neural network is described. Artificial datasets are generated, and the efficacy of the proposed method for both regression and classification tasks is demonstrated. The results obtained for the regression task indicated that the proposed method is capable of obtaining analytical quality derivatives, and in the case of the classification task, the least relevant features could be identified.
-
Reliable probability estimation is of crucial importance in many real-world applications where there is inherent (aleatoric) uncertainty. Probability-estimation models are trained on observed outcomes (e.g. whether it has rained or not, or whether a patient has died or not), because the ground-truth probabilities of the events of interest are typically unknown. The problem is therefore analogous to binary classification, with the difference that the objective is to estimate probabilities rather than predicting the specific outcome. This work investigates probability estimation from high-dimensional data using deep neural networks. There exist several methods to improve the probabilities generated by these models but they mostly focus on model (epistemic) uncertainty. For problems with inherent uncertainty, it is challenging to evaluate performance without access to ground-truth probabilities. To address this, we build a synthetic dataset to study and compare different computable metrics. We evaluate existing methods on the synthetic data as well as on three real-world probability estimation tasks, all of which involve inherent uncertainty: precipitation forecasting from radar images, predicting cancer patient survival from histopathology images, and predicting car crashes from dashcam videos. We also give a theoretical analysis of a model for high-dimensional probability estimation which reproduces several of the phenomena evinced in our experiments. Finally, we propose a new method for probability estimation using neural networks, which modifies the training process to promote output probabilities that are consistent with empirical probabilities computed from the data. The method outperforms existing approaches on most metrics on the simulated as well as real-world data.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1017/eds.2023.7
