Information FOMO: The unhealthy fear of missing out on information. A method for removing misleading data for healthier models
Not on the map. Why?
Abstract (arXiv)
Misleading or unnecessary data can have out-sized impacts on the health or accuracy of Machine Learning (ML) models. We present a Bayesian sequential selection method, akin to Bayesian experimental design, that identifies critically important information within a dataset, while ignoring data that is either misleading or brings unnecessary complexity to the surrogate model of choice. Our method improves sample-wise error convergence and eliminates instances where more data leads to worse performance and instabilities of the surrogate model, often termed sample-wise ``double descent''. We find these instabilities are a result of the complexity of the underlying map and linked to extreme events and heavy tails. Our approach has two key features. First, the selection algorithm dynamically couples the chosen model and data. Data is chosen based on its merits towards improving the selected model, rather than being compared strictly against other data. Second, a natural convergence of the method removes the need for dividing the data into training, testing, and validation sets. Instead, the selection metric inherently assesses testing and validation error through global statistics of the model. This ensures that key information is never wasted in testing or validation. The method is applied using both Gaussian process regression and deep neural network surrogate models.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 26 of its 54 references that are in the map. First posted on arXiv 27 Aug 2022.
Builds on
26 of 54 references.
References · 26
- 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
- 2019DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operatorscited by 2,018
- 2018Reconciling modern machine learning practice and the bias-variance trade-offcited by 927
- 2019Deep Double Descent: Where Bigger Models and More Data Hurtcited by 672
- 2019Surprises in High-Dimensional Ridgeless Least Squares Interpolationcited by 513
- 2020Bayesian Deep Learning and a Probabilistic Perspective of Generalizationcited by 480
- 2019The generalization error of random features regression: Precise asymptotics and double descent curvecited by 413
- 2018Not All Samples Are Created Equal: Deep Learning with Importance Samplingcited by 328
- 2017High-dimensional dynamics of generalization error in neural networkscited by 241
- 2013Stochastic Gradient Descent, Weighted Sampling, and the Randomized Kaczmarz algorithmcited by 133
- 2020Optimal Regularization Can Mitigate Double Descentcited by 109
- 2014Stochastic Optimization with Importance Samplingcited by 91
14 more
- 2015Variance Reduction in SGD by Distributed Importance Samplingcited by 79
- 2019Data-dependent Sample Complexity of Deep Neural Networks via Lipschitz Augmentationcited by 71
- 2018A jamming transition from under- to over-parametrization affects loss landscape and generalizationcited by 67
- 2018The jamming transition as a paradigm to understand the loss landscape of deep neural networkscited by 62
- 2019Improved Sample Complexities for Deep Networks and Robust Classification via an All-Layer Margincited by 56
- 2020Early Stopping in Deep Networks: Double Descent and How to Eliminate itcited by 41
- 2019More Data Can Hurt for Linear Regression: Sample-wise Double Descentcited by 37
- 2022Discovering and forecasting extreme events via active learning in neural operatorscited by 32
- 2015Even Faster Accelerated Coordinate Descent Using Non-Uniform Samplingcited by 23
- 2017Safe Adaptive Importance Samplingcited by 19
- 2019Output-weighted optimal sampling for Bayesian regression and rare event statistics using few samplescited by 7
- 2015Stochastic Dual Coordinate Ascent with Adaptive Probabilitiescited by 5
- 2017Faster Coordinate Descent via Adaptive Importance Samplingcited by 5
- 2022Structure and Distribution Metric for Quantifying the Quality of Uncertainty: Assessing Gaussian Processes, Deep Neural Nets, and Deep Neural Operators for Regressioncited by 5
Built on by
Citing papers per year: 2023: 1.
1 paper.