Fast Forward Selection to Speed Up Sparse Gaussian Process Regression

By Matthias W. Seeger, Christopher K. I. Williams, Neil D. Lawrence. 2003

519 citations recorded by the source corpus.

Subfields: Optimization & Training, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Optimization & Training

In Probabilistic & Bayesian Methods

Citations in the corpus

Taking the Human Out of the Loop: A Review of Bayesian Optimization (2016, methodology): “1) Sparse pseudo-input Gaussian processes (SPGP): One early approach to modelling large n with Gaussian processes considered using m n inducing pseudo-inputs to reduce the rank of the covariance matrix to m , resulting in a signiﬁcant reduction in computational cost [137], [140].”
Information-theoretic Online Memory Selection for Continual Learning (2022, neutral): “For instance, Seeger et al. (2003) adopt the information gain to select inducing points for a scalable Gaussian process method.”
Functional Regularisation for Continual Learning using Gaussian Processes (2019, methodology): “…processes (Rasmussen & Williams, 2005), and makes use of the rich literature on sparse GP methods (Csato & Opper, 2002; Lawrence et al., 2002; Seeger et al., 2003; Quiñonero-Candela & Rasmussen, 2005; Snelson & Ghahramani, 2006; Titsias, 2009; Hensman et al., 2013; Bui et al., 2017b), which…”
Gaussian Processes for Big Data (2013, background): “An alternative is to build a low rank approximation to the covariance matrix based around ‘inducing vari-ables’ [see e.g. Csat´o and Opper, 2002, Seeger et al., 2003, Qui˜nonero Candela and Rasmussen, 2005, Tit-sias, 2009].”
Gaussian Process Regression Networks (2011, background): “…Jay, 1996; Ver Hoef and Barry, 1998; Boyle and Frean, 2004), 6) standard independent Gaussian processes (GP), 7) and the DTC (Csat´o and Opper, 2001; Seeger et al., 2003; Qui˜nonero-Candela and Rasmussen, 2005; Rasmussen and Williams, 2006) , 8) FITC (Snelson and Ghahramani, 2006), and 9) PITC…”
Gaussian Processes for Regression and Optimisation (2007, background): “[71, 80, 13, 14, 15, 66]), including the reduced rank approximation of Quiñonero-Candela and Rasmussen [55] to be considered in chapter 5.”
Variable Noise and Dimensionality Reduction for Sparse Gaussian processes (2006, methodology): “There have been many methods proposed in recent years to address this problem, and bring the scaling down to O ( M 2 N ) where M N [Tresp, 2000, Smola and Bartlett, 2001, Williams and Seeger, 2001, Csat´o, 2002a,b, Lawrence et al., 2003, Seeger et al., 2003, Seeger, 2003, Qui˜nonero Candela, 2004].”
Variational Learning of Inducing Variables in Sparse Gaussian Processes (2009, methodology): “A related objective function used in (Seeger et al., 2003) corresponds to the projected process approximation (PP).”
Computationally Efficient Convolved Multiple Output Gaussian Processes (2011, methodology): “3 DETERMINISTIC LIKELIHOOD In Quiñonero-Candela and Rasmussen (2005), the relationship between the projected process approximation (Csató and Opper, 2001; Seeger et al., 2003) and the FI(T)C and PI(T)C approximations is elucidated.”
Accelerating evolutionary algorithms with Gaussian process fitness function models (2005, background): “These could be addressed by a sparse GP implementation [26].”
Bayesian Gaussian Process Latent Variable Model (2010, methodology): “where q(f ∗ |x∗) is a factorized Gaussian distribution where each factor takes the form of the projected process predictive distribution (Csató and Opper, 2002; Seeger et al., 2003; Rasmussen and Williams, 2006).”
Efficient Multioutput Gaussian Processes through Variational Inducing Kernels (2010, background): “The form of the bound is very similar to the projected process approximation, also known as DTC (Csató and Opper, 2001; Seeger et al., 2003; Rasmussen and Williams, 2006).”
GP-BayesFilters: Bayesian filtering using Gaussian process prediction and observation models (2008, methodology): “This optimization is smoother than that done in other sparse GP work such as (Csató and Opper 2002; Seeger and Williams 2003; Smola and Bartlett 2001) , where active points are a subset of the training points.”
Information Theory, Inference, and Learning Algorithms (2004, background): “Methods for reducing the complexity of data modelling with Gaussian processes remain an active research area (Poggio and Girosi, 1990; Luo and Wahba, 1997; Tresp, 2000; Williams and Seeger, 2001; Smola and Bartlett, 2001; Rasmussen, 2002; Seeger et al., 2003; Opper and Winther, 2000).”