Maximin effects in inhomogeneous large-scale data
From MaRDI portal
Abstract: Large-scale data are often characterized by some degree of inhomogeneity as data are either recorded in different time regimes or taken from multiple sources. We look at regression models and the effect of randomly changing coefficients, where the change is either smoothly in time or some other dimension or even without any such structure. Fitting varying-coefficient models or mixture models can be appropriate solutions but are computationally very demanding and often return more information than necessary. If we just ask for a model estimator that shows good predictive properties for all regimes of the data, then we are aiming for a simple linear model that is reliable for all possible subsets of the data. We propose the concept of "maximin effects" and a suitable estimator and look at its prediction accuracy from a theoretical point of view in a mixture model with known or unknown group structure. Under certain circumstances the estimator can be computed orders of magnitudes faster than standard penalized regression estimators, making computations on large-scale data feasible. Empirical examples complement the novel methodology and theory.
Recommendations
- Confidence Intervals for Maximin Effects in Inhomogeneous Large-Scale Data
- Partial least squares for heterogeneous data
- Randomized maximum-contrast selection: subagging for large-scale regression
- A partially linear framework for massive heterogeneous data
- Estimation and inference for very large linear mixed effects models
Cites work
- \(\ell_{1}\)-penalization for mixture regression models
- A comparison of the Lasso and marginal regression
- Asymptotically efficient adaptive allocation rules
- Atomic decomposition by basis pursuit
- Classifier technology and the illusion of progress
- Efficient Estimation and Inferences for Varying-Coefficient Models
- Estimating time-varying networks
- Finite mixture models
- Finite-time analysis of the multiarmed bandit problem
- scientific article; zbMATH DE number 3921733 (Why is no real title available?)
- scientific article; zbMATH DE number 3954047 (Why is no real title available?)
- scientific article; zbMATH DE number 3567782 (Why is no real title available?)
- scientific article; zbMATH DE number 472973 (Why is no real title available?)
- scientific article; zbMATH DE number 2006097 (Why is no real title available?)
- scientific article; zbMATH DE number 1487502 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- Linear Minimax Regret Estimation of Deterministic Parameters with Bounded Data Uncertainties
- Prediction, Learning, and Games
- Regret in online combinatorial optimization
- Regret in the on-line decision problem
- Restricted eigenvalue properties for correlated Gaussian designs
- Ridge Regression: Biased Estimation for Nonorthogonal Problems
- Robust Estimation of a Location Parameter
- Robust regression: Asymptotics, conjectures and Monte Carlo
- Sparse recovery under weak moment assumptions
- Statistical decision functions which minimize the maximum risk
- Statistical estimation in varying coefficient models
- Statistics for high-dimensional data. Methods, theory and applications.
Cited in
(22)- Statistics for big data: a perspective
- Wide consensus aggregation in the Wasserstein space. Application to location-scatter families
- Comments on ``Data science, big data and statistics
- Confidence Intervals for Maximin Effects in Inhomogeneous Large-Scale Data
- A partially linear framework for massive heterogeneous data
- Partial least squares for heterogeneous data
- Comments on: ``A random forest guided tour
- High-Dimensional Factor Regression for Heterogeneous Subpopulations
- Distributionally robust and generalizable inference
- Statistical Inference for Maximin Effects: Identifying Stable Associations across Multiple Studies
- Non parametric maximin aggregation for data with inhomogeneity
- Multi-Dimensional Domain Generalization with Low-Rank Structures
- Optimal Multitask Linear Regression and Contextual Bandits under Sparse Heterogeneity
- Causality pursuit from heterogeneous environments via neural adversarial invariance learning
- Robust Inference for Federated Meta-Learning
- Robust angle-based transfer learning in high dimensions
- Engression: extrapolation through the Lens of distributional regression
- Invariant subspace decomposition
- Joint and Individual Component Regression
- Distributionally robust learning for multisource unsupervised domain adaptation
- Causality-Oriented Robustness: Exploiting General Noise Interventions
- Learning models with uniform performance via distributionally robust optimization
This page was built for publication: Maximin effects in inhomogeneous large-scale data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2515497)
