Scalable Variational Gaussian Process Classification

Abstract

Gaussian process classification is a popular method with a number of appealing properties. We show how to scale the model within a variational inducing point framework, outperforming the state of the art on benchmark datasets. Importantly, the variational formulation can be exploited to allow classification in problems with millions of data points, as we demonstrate in experiments.

16 pages, 9 figures

References in corpus (6)

• Scikit-learn: Machine Learning in Python
• ADADELTA: An Adaptive Learning Rate Method
• Gaussian Processes for Big Data
• Gaussian Process Regression Networks
• Distributed Variational Inference in Sparse Gaussian Process Regression and Latent Variable Models
• Gaussian Process Models with Parallelization and GPU acceleration

Cited by in corpus (106)

• Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift
• Neural network an1alysis of sleep stages enables efficient diagnosis of narcolepsy
• When Gaussian Process Meets Big Data: A Review of Scalable GPs
• Functional Regularisation for Continual Learning with Gaussian Processes
• Random Feature Expansions for Deep Gaussian Processes
• A Framework for Interdomain and Multioutput Gaussian Processes
• Understanding Probabilistic Sparse Gaussian Process Approximations
• Multi-fidelity classification using Gaussian processes: accelerating the prediction of large-scale computational models
• On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty
• MCMC for Variationally Sparse Gaussian Processes
• Heterogeneous Multi-output Gaussian Process Prediction
• Rates of Convergence for Sparse Variational Gaussian Process Regression
• A Survey of Optimization Methods from a Machine Learning Perspective
• Variational Bayes survival analysis for unemployment modelling
• Exact Gaussian Processes on a Million Data Points
• Kernel methods through the roof: handling billions of points efficiently
• A Survey of Bayesian Statistical Approaches for Big Data
• Interactive Imitation Learning of Bimanual Movement Primitives
• Multi-class Gaussian Process Classification with Noisy Inputs
• Early Detection of COVID-19 Hotspots Using Spatio-Temporal Data
• Evaluating uncertainties in electrochemical impedance spectra of solid oxide fuel cells
• Tuning Fairness by Balancing Target Labels
• Convergence of Sparse Variational Inference in Gaussian Processes Regression
• Introducing instance label correlation in multiple instance learning. Application to cancer detection on histopathological images
• GP-Tree: A Gaussian Process Classifier for Few-Shot Incremental Learning
• Quantifying Point-Prediction Uncertainty in Neural Networks via Residual Estimation with an I/O Kernel
• Fast Matrix Square Roots with Applications to Gaussian Processes and Bayesian Optimization
• Personalized Federated Learning with Gaussian Processes
• A Tutorial on Sparse Gaussian Processes and Variational Inference
• Improving predictions of Bayesian neural nets via local linearization
• Global Stochastic Optimization of Stellarator Coil Configurations
• Transforming Gaussian Processes With Normalizing Flows
• Accelerating Monte Carlo Bayesian Inference via Approximating Predictive Uncertainty over Simplex
• Scaling Gaussian Processes with Derivative Information Using Variational Inference
• Distribution Calibration for Regression
• A Global-Local Approximation Framework for Large-Scale Gaussian Process Modeling
• Causal-BALD: Deep Bayesian Active Learning of Outcomes to Infer Treatment-Effects from Observational Data
• Learning and Interpreting Multi-Multi-Instance Learning Networks
• Parametric Gaussian Process Regressors
• Continual Multi-task Gaussian Processes
• Multi-Class Gaussian Process Classification Made Conjugate: Efficient Inference via Data Augmentation
• A statistical theory of cold posteriors in deep neural networks
• Short-Term Solar Irradiance Forecasting Using Calibrated Probabilistic Models
• Fidelity-Weighted Learning
• Surrogate-based optimization of system architectures subject to hidden constraints
• Probabilistic selection of inducing points in sparse Gaussian processes
• Variational Orthogonal Features
• Multi-task Learning for Aggregated Data using Gaussian Processes
• Sparse Orthogonal Variational Inference for Gaussian Processes
• Interpretable Dynamics Models for Data-Efficient Reinforcement Learning
• Pathwise Conditioning of Gaussian Processes
• Spatio-Temporal Variational Gaussian Processes
• Modular Gaussian Processes for Transfer Learning
• Web Mining to Inform Locations of Charging Stations for Electric Vehicles
• Bayesian Image Classification with Deep Convolutional Gaussian Processes
• Sparse Spectrum Warped Input Measures for Nonstationary Kernel Learning
• Variational Autoencoding of PDE Inverse Problems
• Deep Mixed Effect Model using Gaussian Processes: A Personalized and Reliable Prediction for Healthcare
• Quadruply Stochastic Gaussian Processes
• Scalable Bayesian dynamic covariance modeling with variational Wishart and inverse Wishart processes
• An Infinite-Feature Extension for Bayesian ReLU Nets That Fixes Their Asymptotic Overconfidence
• Ranking Creative Language Characteristics in Small Data Scenarios
• Non-Parametric Calibration for Classification
• SKIing on Simplices: Kernel Interpolation on the Permutohedral Lattice for Scalable Gaussian Processes
• Recent Advances in Data-Driven Wireless Communication Using Gaussian Processes: A Comprehensive Survey
• Scalable Gaussian Process Inference with Finite-data Mean and Variance Guarantees
• Kernelized Capsule Networks
• Scalable Cross Validation Losses for Gaussian Process Models
• GPIRT: A Gaussian Process Model for Item Response Theory
• Variational Auto-Regressive Gaussian Processes for Continual Learning
• Multi-output Gaussian Processes for Uncertainty-aware Recommender Systems
• On Signal-to-Noise Ratio Issues in Variational Inference for Deep Gaussian Processes
• Inter-domain Deep Gaussian Processes
• Scalable Gaussian Process Classification with Additive Noise for Various Likelihoods
• Uncertainty-aware Remaining Useful Life predictor
• An innovative adaptive kriging approach for efficient binary classification of mechanical problems
• Compositional Modeling of Nonlinear Dynamical Systems with ODE-based Random Features
• Knot Selection in Sparse Gaussian Processes
• Automated Augmented Conjugate Inference for Non-conjugate Gaussian Process Models
• Multi-resolution spatio-temporal prediction with application to wind power generation
• Input Dependent Sparse Gaussian Processes
• Deep Neural Networks as Point Estimates for Deep Gaussian Processes
• Variational bridge constructs for approximate Gaussian process regression
• Bayes-Newton Methods for Approximate Bayesian Inference with PSD Guarantees
• Quantifying Predictive Uncertainty in Medical Image Analysis with Deep Kernel Learning
• Scalable Variational Gaussian Processes via Harmonic Kernel Decomposition
• Gaussian Processes with Differential Privacy
• Generative Particle Variational Inference via Estimation of Functional Gradients
• Sparse within Sparse Gaussian Processes using Neighbor Information
• Active Assessment of Prediction Services as Accuracy Surface Over Attribute Combinations
• Estimation of Bivariate Structural Causal Models by Variational Gaussian Process Regression Under Likelihoods Parametrised by Normalising Flows
• Towards Credit-Fraud Detection via Sparsely Varying Gaussian Approximations
• Self-explaining variational posterior distributions for Gaussian Process models
• Large-Scale Learning with Fourier Features and Tensor Decompositions
• Direct loss minimization algorithms for sparse Gaussian processes
• Bayesian Meta-Learning Through Variational Gaussian Processes
• Quantile Propagation for Wasserstein-Approximate Gaussian Processes
• Validating Gaussian Process Models with Simulation-Based Calibration
• Healing Products of Gaussian Processes
• Mixtures of Gaussian Processes for regression under multiple prior distributions
• Orthogonally Decoupled Variational Fourier Features
• Recyclable Gaussian Processes
• Implicit Priors for Knowledge Sharing in Bayesian Neural Networks
• Bayesian task embedding for few-shot Bayesian optimization
• Scalable GAM using sparse variational Gaussian processes
• Efficient Deep Gaussian Process Models for Variable-Sized Input