Kernels for vector-valued functions: a review

DOI10.1561/2200000036MaRDI QIDQ2903301zbMATH OpenOpenAlexWikidataFDO

Authors Mauricio A. Álvarez, Lorenzo Rosasco, Neil D. Lawrence

Publication date 8 August 2012

Published in Foundations and Trends in Machine Learning (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1106.6251, http://eprints.whiterose.ac.uk/114503/1/1106.6251v2.pdf

Mathematics Subject Classification ID

Gaussian processes (60G15) Learning and adaptive systems in artificial intelligence (68T05)

Abstract: Kernel methods are among the most popular techniques in machine learning. From a frequentist/discriminative perspective they play a central role in regularization theory as they provide a natural choice for the hypotheses space and the regularization functional through the notion of reproducing kernel Hilbert spaces. From a Bayesian/generative perspective they are the key in the context of Gaussian processes, where the kernel function is also known as the covariance function. Traditionally, kernel methods have been used in supervised learning problem with scalar outputs and indeed there has been a considerable amount of work devoted to designing and learning kernels. More recently there has been an increasing interest in methods that deal with multiple outputs, motivated partly by frameworks like multitask learning. In this paper, we review different methods to design or learn valid kernel functions for multiple outputs, paying particular attention to the connection between probabilistic and functional methods.

Recommendations

• On Learning Vector-Valued Functions
• Kernel methods in machine learning
• Multi-output learning via spectral filtering
• Operator-valued kernels for learning from functional response data
• scientific article; zbMATH DE number 5968915

Cited in

(only showing first 100 items - show all)

• A Bayesian optimization approach to find Nash equilibria
• Output Fisher embedding regression
• Ensembles for multi-target regression with random output selections
• Multi-target regression via input space expansion: treating targets as inputs
• Reduced order models for many-query subsurface flow applications
• Varying-coefficient models for geospatial transfer learning
• Multi-task learning in vector-valued reproducing kernel Banach spaces with the \(\ell^1\) norm
• Large scale multi-label learning using Gaussian processes
• Bayesian optimization of functional output in inverse problems
• Multi-fidelity regression using artificial neural networks: efficient approximation of parameter-dependent output quantities
• Mixture of multivariate Gaussian processes for classification of irregularly sampled satellite image time-series
• Do ideas have shape? Idea registration as the continuous limit of artificial neural networks
• Calibrate, emulate, sample
• Deep coregionalization for the emulation of simulation-based spatial-temporal fields
• Phase retrieval of complex and vector-valued functions
• A non-intrusive multifidelity method for the reduced order modeling of nonlinear problems
• Dynamic mode decomposition in vector-valued reproducing kernel Hilbert spaces for extracting dynamical structure among observables
• Just interpolate: kernel ``ridgeless regression can generalize
• High-dimensional Bayesian optimization using low-dimensional feature spaces
• Multi-target prediction: a unifying view on problems and methods
• A unifying representer theorem for inverse problems and machine learning
• Bayesian optimization of variable-size design space problems
• Data-driven model order reduction for problems with parameter-dependent jump-discontinuities
• Residual Gaussian process: a tractable nonparametric Bayesian emulator for multi-fidelity simulations
• Operator learning approach for the limited view problem in photoacoustic tomography
• Symmetry exploits for Bayesian cubature methods
• A Riemannian gossip approach to subspace learning on Grassmann manifold
• Grouped Gaussian processes for solar power prediction
• Multivariate versus univariate Kriging metamodels for multi-response simulation models
• A constrained matrix-variate Gaussian process for transposable data
• Cross-covariance functions for multivariate geostatistics
• Convex optimization in sums of Banach spaces
• Multi-fidelity surrogate modeling using long short-term memory networks
• Machine learning-based multi-objective optimization for efficient identification of crystal plasticity model parameters
• GParareal: a time-parallel ODE solver using Gaussian process emulation
• Neuronal spike train analysis using Gaussian process models
• Input output kernel regression: supervised and semi-supervised structured output prediction with operator-valued kernels
• Kernel-based methods for vector-valued data with correlated components
• Experimental design for nonparametric correction of misspecified dynamical models
• Convergence rates for matrix P-greedy variants
• Multi-output learning via spectral filtering
• Exact Bayesian Inference in Spatiotemporal Cox Processes Driven by Multivariate Gaussian Processes
• Operator-valued kernel-based vector autoregressive models for network inference
• Generalized probabilistic principal component analysis of correlated data
• Symplectic Gaussian process regression of maps in Hamiltonian systems
• scientific article; zbMATH DE number 7370541 (Why is no real title available?)
• scientific article; zbMATH DE number 7625160 (Why is no real title available?)
• Two-Layer Neural Networks with Values in a Banach Space
• A Comparative Study of Pairwise Learning Methods Based on Kernel Ridge Regression
• A representer theorem for deep neural networks
• Bayesian optimization for policy search via online-offline experimentation
• Computationally efficient convolved multiple output Gaussian processes
• On Learning Vector-Valued Functions
• Surrogate Modeling with Gaussian Processes for an Inverse Problem in Polymer Dynamics
• Deep Gaussian process for multi-objective Bayesian optimization
• Bayesian optimization with safety constraints: safe and automatic parameter tuning in robotics
• Intercorrelated random fields with bounds and the Bayesian identification of their parameters: Application to linear elastic struts and fibers
• Structured learning of rigid‐body dynamics: A survey and unified view from a robotics perspective
• Large scale multi-output multi-class classification using Gaussian processes
• On Negative Transfer and Structure of Latent Functions in Multioutput Gaussian Processes
• Gaussian process hydrodynamics
• Learning system parameters from Turing patterns
• A Bayesian decision framework for optimizing sequential combination antiretroviral therapy in people with HIV
• Hilbert C∗-Module for Analyzing Structured Data
• Kernel methods are competitive for operator learning
• Learning nonparametric ordinary differential equations from noisy data
• Interpolation with uncoupled separable matrix-valued kernels
• Bayesian estimation of large-scale simulation models with Gaussian process regression surrogates
• Gaussian processes for Bayesian inverse problems associated with linear partial differential equations
• Uncertainty modeling and propagation for groundwater flow: a comparative study of surrogates
• Heteroscedastic Gaussian process regression for material structure-property relationship modeling
• Nonparametric Modeling and Prognosis of Condition Monitoring Signals Using Multivariate Gaussian Convolution Processes
• A review on statistical and machine learning competing risks methods
• Gaussian kernel with correlated variables for incomplete data
• Bayesian optimisation for constrained problems
• GFN: a graph feedforward network for resolution-invariant reduced operator learning in multifidelity applications
• Gradient-enhanced deep Gaussian processes for multifidelity modeling
• From kernel methods to neural networks: a unifying variational formulation
• Operator learning with Gaussian processes
• Targeted adaptive design
• Intrinsic estimation of Frenet curvatures for 3D trajectory analysis
• On the continuity and differentiability of matrix-valued covariance kernels depending on Euclidean or spherical distances
• Regularized reduced-rank regression for structured output prediction
• Wrapped Gaussian process functional regression model for batch data on Riemannian manifolds
• Scalable spatiotemporally varying coefficient modeling with Bayesian kernelized tensor regression
• Efficient training of Gaussian processes with tensor product structure
• A joint kriging model with application to constrained classification
• Bayesian optimization
• Deep latent force models: ODE-based process convolutions for Bayesian deep learning
• Diffeomorphic measure matching with kernels for generative modeling
• Dependence in constrained Bayesian optimization. When do we need it and how does it help?
• Additive multi-task learning models and task diagnostics
• Bayesian learning with Gaussian processes for low-dimensional representations of time-dependent nonlinear systems
• Multi-fidelity reduced-order surrogate modelling
• Estimating individual treatment effects using non-parametric regression models: a review
• Improving the lag window estimators of the spectrum and memory for long-memory stationary Gaussian processes
• Towards empirical process theory for vector-valued functions: metric entropy of smooth function classes
• A characterization of multioutput learnability
• Consistency of some sequential experimental design strategies for excursion set estimation based on vector-valued Gaussian processes
• Aggregated multi-output Gaussian processes with knowledge transfer across domains

This page was built for publication: Kernels for vector-valued functions: a review

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2903301)