Sample-based Uncertainty Quantification with a Single Deterministic Neural Network
Not on the map. Why?
Abstract (arXiv)
Development of an accurate, flexible, and numerically efficient uncertainty quantification (UQ) method is one of fundamental challenges in machine learning. Previously, a UQ method called DISCO Nets has been proposed (Bouchacourt et al., 2016), which trains a neural network by minimizing the energy score. In this method, a random noise vector in ℝ10--100 is concatenated with the original input vector in order to produce a diverse ensemble forecast despite using a single neural network. While this method has shown promising performance on a hand pose estimation task in computer vision, it remained unexplored whether this method works as nicely for regression on tabular data, and how it competes with more recent advanced UQ methods such as NGBoost. In this paper, we propose an improved neural architecture of DISCO Nets that admits faster and more stable training while only using a compact noise vector of dimension ~ O(1). We benchmark this approach on miscellaneous real-world tabular datasets and confirm that it is competitive with or even superior to standard UQ baselines. Moreover we observe that it exhibits better point forecast performance than a neural network of the same size trained with the conventional mean squared error. As another advantage of the proposed method, we show that local feature importance computation methods such as SHAP can be easily applied to any subregion of the predictive distribution. A new elementary proof for the validity of using the energy score to learn predictive distributions is also provided.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 37 of its 62 references that are in the map. First posted on arXiv 17 Sep 2022.
Builds on
37 of 62 references.
References · 37
- 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
- 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
- 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,888
- 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,455
- 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,315
- 2016Deep Networks with Stochastic Depthcited by 973
- 2019Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methodscited by 903
- 2020A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challengescited by 813
- 2021A Survey of Uncertainty in Deep Neural Networkscited by 675
- 2017Self-Normalizing Neural Networkscited by 591
- 2019A Simple Baseline for Bayesian Uncertainty in Deep Learningcited by 555
- 2017Distributional Reinforcement Learning with Quantile Regressioncited by 464
25 more
- 2018Implicit Quantile Networks for Distributional Reinforcement Learningcited by 362
- 2017Snapshot Ensembles: Train 1, get M for freecited by 349
- 2020Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Countscited by 163
- 2020Efficiently Sampling Functions from Gaussian Process Posteriorscited by 125
- 2021DEUP: Direct Epistemic Uncertainty Predictioncited by 102
- 2018Uncertainty in Neural Networks: Approximately Bayesian Ensemblingcited by 94
- 2019NGBoost: Natural Gradient Boosting for Probabilistic Predictioncited by 94
- 2019Fully Parameterized Quantile Function for Distributional Reinforcement Learningcited by 87
- 2020Depth Uncertainty in Neural Networkscited by 56
- 2021Estimating and Evaluating Regression Predictive Uncertainty in Deep Object Detectorscited by 37
- 2021Probabilistic Forecasting with Generative Networks via Scoring Rule Minimizationcited by 35
- 2021Distributional Reinforcement Learning for Multi-Dimensional Reward Functionscited by 21
- 2021Probabilistic Gradient Boosting Machines for Large-Scale Probabilistic Regressioncited by 20
- 2016DISCO Nets: DISsimilarity COefficient Networkscited by 18
- 2020Deep transformation models: Tackling complex regression problems with neural network based transformation modelscited by 18
- 2021A deep generative model for probabilistic energy forecasting in power systems: normalizing flowscited by 18
- 2022Multivariate Probabilistic Forecasting of Intraday Electricity Prices using Normalizing Flowscited by 14
- 2020Sample-based Distributional Policy Gradientcited by 12
- 2021Non-Gaussian Gaussian Processes for Few-Shot Regressioncited by 12
- 2022Instance-Based Uncertainty Estimation for Gradient-Boosted Regression Treescited by 9
- 2018The Exact Equivalence of Distance and Kernel Methods for Hypothesis Testingcited by 7
- 2022Short-Term Density Forecasting of Low-Voltage Load using Bernstein-Polynomial Normalizing Flowscited by 6
- 2022Distributional Gradient Boosting Machinescited by 5
- 2022Conditional Approximate Normalizing Flows for Joint Multi-Step Probabilistic Forecasting with Application to Electricity Demandcited by 2
- 2022A Deep Learning Approach to Probabilistic Forecasting of Weathercited by 1
Built on by
Citing papers per year: 2023: 1; 2024: 1; 2025: 1.
3 papers.