The kernel recursive least-squares algorithm

By Yaakov Engel, Shie Mannor, R. Meir. IEEE Transactions on Signal Processing, 2004

1,084 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/2ab464a74e0bff7ab6e84e6c7d04702548a655de

How researchers see this paper

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

In Computational Mechanics

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

Dlib-ml: A Machine Learning Toolkit (2009, methodology): “Another good example of a generic kernel algorithm provided by the library is the kernel RLS technique introduced by Engel et al. (2004).”
On-line policy learning and adaptation for real-time personalization of an artificial pancreas (2015, methodology): “Further details on IOS can be found in the works of Nguyen-Tuong and Peters (2011a) and Engel et al. (2004), and references therein.”
On-line quantile regression in the RKHS (Reproducing Kernel Hilbert Space) for operational probabilistic forecasting of wind power (2016, background): “This idea is the base of 12 sparsification [37, 38].”
Quantized Kernel Least Mean Square Algorithm (2012, methodology): “These algorithms include the kernel least mean square (KLMS) [17], [18], kernel affine projection algorithms (KAPAs) [19], kernel recursive least squares (KRLS) [20], extended KRLS [21], etc.”
The Kernel Least-Mean-Square Algorithm (2008, background): “As pointed out in [13] and [27], the ill-posedness lies exactly in this inverse operation, and different kinds of regularization techniques can be viewed as different ways of getting meaningful inverse approximations.”
On-line regression algorithms for learning mechanical models of robots: A survey (2011, methodology): “In [69], lgp is augmented with an on-line sparsiﬁcation algorithm inspired from what is used in Kernel Recursive Least-Squares ( krls ) [34].”
Sample-efficient batch reinforcement learning for dialogue management optimization (2011, methodology): “This can be solved using the Engel’s dictionnary approach [Engel et al. 2004].”
A Tutorial on Environment-Aware Communications via Channel Knowledge Map for 6G (2023, methodology): “One popular incremental learning algorithm is the Recursive Least Squares (RLS) algorithm [222], which implements a recursive update formula to update the CKM with new data, making it a computationally efficient and scalable solution.”
Kernel Methods and Machine Learning (2014, neutral): “For example, see [62, 157].”
Online Kernel Principal Component Analysis: A Reduced-Order Model (2012, methodology): “In this paper, we examine a distance-based selection criterion, initially investigated for sparse regression [26], [27], and we show its appropriateness for the online kernel-PCA algorithm.”
Estimation of the forgetting factor in kernel recursive least squares (2012, background): “The standard RLS formulation often includes a forgetting factor λ ∈ (0, 1] that allows it to deal with non-stationary systems to some extent, while in order to obtain true tracking capabilities a more general extended recursive least-squares (EX-RLS) algorithm is required [1, 2].”
A kernel machine-based secure data sensing and fusion scheme in wireless sensor networks for the cyber-physical systems (2016, methodology): “Moreover, in order to reduce the computational complexity and improve the accuracy of the prediction algorithm with small number of data items, the proposed scheme employs GM to obtain the initial predicted value, then with the help of KRLS learning algorithm [26], our scheme makes the predicted val”
Decentralized Online Learning With Kernels (2017, methodology): “These combine functional extensions of stochastic gradient method with compressions of the function parameterization independently of the optimization problem to which they are applied [20]–[24] or approximate the kernel during training [25]–[29], and at best converge on average.”