Novel kernelized LSTD method improves Q-function approximation in RL.
problem Improving policy evaluation in reinforcement learning.
method Manifold regularization applied to kernelized LSTD.
result Superior performance in Q-function approximation compared to existing methods.
New algorithm LSTD(λ)-RP uses random projections and eligibility traces for efficient reinforcement learning.
problem Policy evaluation in high-dimensional feature spaces with linear function approximation.
method Proposes LSTD(λ)-RP algorithm combining random projections and eligibility traces. result Demonstrates improved performance and better error bounds compared to prior methods.
In this analytical study we derive the optimal unbiased value estimator (MVU) and compare its statistical risk to three well known value estimators: Temporal Difference learning (TD), Monte Carlo estimation (MC) and Least-Squares Temporal Difference Learning (LSTD). We demonstrate that LSTD is equivalent to the MVU if …
This paper presents four different ways of looking at the well-known Least Squares Temporal Differences (LSTD) algorithm for computing the value function of a Markov Reward Process, each of them leading to different insights: the operator-theory approach via the Galerkin method, the statistical approach via instrumenta…
Kernel-based methods improve policy evaluation in MRP models.
problem Estimating value functions in infinite-horizon discounted MRP models.
method Kernel-based temporal difference methods using reproducing kernel Hilbert spaces.
result Optimal error bounds derived for the kernel-based LSTD estimate.
We propose a stochastic approximation (SA) based method with randomization of samples for policy evaluation using the least squares temporal difference (LSTD) algorithm. Our proposed scheme is equivalent to running regular temporal difference learning with linear function approximation, albeit with samples picked unifo…
LSTD is a popular algorithm for value function approximation. Whenever the number of features is larger than the number of samples, it must be paired with some form of regularization. In particular, L1-regularization methods tend to perform feature selection by promoting sparsity, and thus, are well-suited for high-dim…
Study LSTD on LQR, finds sample complexity for value function estimation.
problem Sample complexity of RL on continuous problems.
method Least-Squares Temporal Difference (LSTD) on Linear Quadratic Regulator (LQR).
result First finite-time analysis of LQR value function estimation.
KBB algorithm reduces sample complexity for policy evaluation in general state spaces.
problem Policy evaluation in large state spaces with high sample complexity.
method Alternates between fitting Bellman residual and estimating value function via adaptive feature set growth.
result Super-linear convergence rates demonstrated, with reductions in sample complexity.
The paper analyzes how over-parameterization affects reinforcement learning performance.
problem Understanding the impact of over-parameterization in reinforcement learning.
method Theoretical analysis of Least-Square Temporal Difference (LSTD) algorithm with random features and asymptotic regime.
result Identification of a double descent phenomenon in reinforcement learning performance.
The paper improves reinforcement learning stability and efficiency with a new theoretical framework.
problem Stability and efficiency in reinforcement learning, especially in data-scarce scenarios.
method Theoretical framework using resampled U- and V-statistics to model experience replay, applied to policy evaluation and kernel ridge regression. result Significant improvements in stability and efficiency, particularly in data-scarce scenarios.
This work characterizes conditions for offline policy evaluation in reinforcement learning.
problem Understanding when classical methods succeed in offline policy evaluation for linear function approximation.
method Control-theoretic and linear-algebraic conditions for classical methods (FQI and LSTD).
result A precise hierarchy of regimes under which these estimators succeed, and a complete picture of their behavior.
In this paper we extend temporal difference policy evaluation algorithms to performance criteria that include the variance of the cumulative reward. Such criteria are useful for risk management, and are important in domains such as finance and process control. We propose both TD(0) and LSTD(lambda) variants with linear…
New algorithms for model selection in off-policy evaluation of reinforcement learning.
problem Hyperparameter tuning for off-policy evaluation methods in reinforcement learning.
method Developed new model-free and model-based selectors with theoretical guarantees and a new experimental protocol.
result New model-free selector, LSTD-Tournament, demonstrates promising empirical performance.
Temporal Difference learning or TD(λ) is a fundamental algorithm in the field of reinforcement learning. However, setting TD's λ parameter, which controls the timescale of TD updates, is generally left up to the practitioner. We formalize the λ selection problem as a bias-variance trade-off where the solution is …
The paper analyzes GTD algorithms with finite-sample bounds.
problem Convergence rate analysis of GTD family of algorithms.
method Formulated as stochastic gradient algorithms and analyzed using saddle-point error.
result Obtained finite-sample bounds on GTD performance.
Study shows model-based methods require fewer samples than model-free methods for LQR tasks.
problem Comparing model-based and model-free methods in reinforcement learning for continuous control tasks.
method An asymptotic analysis of sample complexity for policy evaluation in LQR tasks.
result Model-based methods require asymptotically less samples than model-free methods for policy evaluation in LQR tasks.
Unified framework for PE and TD methods in continuous time and space.
problem Policy evaluation and TD learning in continuous settings.
method Martingale characterization for designing PE algorithms.
result Convergent time-discretized algorithms converge to continuous-time counterparts.
Many unsupervised kernel methods rely on the estimation of the kernel covariance operator (kernel CO) or kernel cross-covariance operator (kernel CCO). Both kernel CO and kernel CCO are sensitive to contaminated data, even when bounded positive definite kernels are used. To the best of our knowledge, there are few well…
Survey of kernels, RKHS, and their applications in machine learning.
problem Understanding kernels and their applications in machine learning.
method Review of historical context, mathematical definitions, and practical applications of kernels.
result Comprehensive overview of kernels, RKHS, and their applications.
To the best of our knowledge, there are no general well-founded robust methods for statistical unsupervised learning. Most of the unsupervised methods explicitly or implicitly depend on the kernel covariance operator (kernel CO) or kernel cross-covariance operator (kernel CCO). They are sensitive to contaminated data, …
Deep neural kernels and Laplace kernel have equivalent RKHS on spheres.
problem Comparing RKHS of deep neural tangent and Laplace kernels.
method Proof of RKHS equivalence using sphere restrictions and kernel properties.
result RKHS of deep neural tangent kernel and Laplace kernel are the same on Sd−1. Study on expressive power of Euclidean kernels and efficient kernel learning.
problem Limiting the expressive power of kernel methods and improving kernel learning efficiency.
method Define Euclidean kernels, analyze their geometric and spectral properties, and develop efficient algorithms for kernel learning.
result Prove limitations on the expressive power of Euclidean kernels and derive efficient algorithms for kernel learning.
Kernel methods linked to feature subspaces and maximal correlation kernels.
problem Understanding kernel methods and their relationship to feature extraction.
method Established a correspondence between feature subspaces and kernels, introduced maximal correlation kernels, and demonstrated their optimality.
result Kernel SVM on maximal correlation kernel achieves minimum prediction error.
Proposes a method to learn a low-rank kernel matrix for graph-based clustering.
problem Challenges in learning an optimal kernel matrix for graph-based clustering.
method Unified framework for graph construction and kernel learning, focusing on a low-rank kernel matrix.
result Efficacy of the proposed method validated through extensive experiments.
Adapts manifold structure for better clustering performance.
problem Lack of consideration for local manifold structure in existing multiple kernel k-means methods.
method Adopts manifold adaptive kernel to integrate local manifold structure of kernels.
result Proposed method outperforms state-of-the-art methods.
PGF kernels analyze spherical data using generalized RBF kernels.
problem Analysis of spherical data.
method Introduced PGF kernels and a semi-parametric learning algorithm.
result PGF kernels generalize RBF kernels for spherical data.
Optimal kernel in KR can be data-dependent, improving model performance.
problem Fixed kernel in KR limits model performance.
method Considered data-dependent kernels for KR, using posterior covariance.
result Data-dependent kernel choice leads to optimal performance.
Quantum kernels can be efficiently embedded into classical feature spaces.
problem Can all quantum kernels be efficiently embedded into classical feature spaces?
method Invoking computational universality and using techniques like random Fourier features, the authors show that certain classes of quantum kernels can be efficiently embedded.
result For shift-invariant and composition kernels, embedding quantum kernels are universal and efficient.
New random feature maps for Laplacian and related kernels.
problem Challenges in approximating the Laplacian kernel and its generalizations.
method Developed random feature maps for Laplacian and related kernels, providing efficient sampling schemes.
result Demonstrated the efficacy of these random feature maps on real datasets.
New method for learning with non-Euclidean data using decomposable kernels.
problem Difficulty in using classical kernels for non-Euclidean data.
method Reproducing kernel Krein space (RKKS) methods for kernels that admit a positive decomposition.
result Invariant kernels can be used for learning in non-Euclidean spaces.
New estimator reduces kernel mean estimation error.
problem Kernel mean estimation in reproducing kernel Hilbert spaces.
method Corrupt data with known distributions and estimate kernel mean under the corrupted distribution.
result The marginalized kernel mean estimator achieves lower estimation error.
A new kernel, Isolation Kernel, simplifies large scale online kernel learning without sacrificing accuracy.
problem Building efficient and scalable kernel-based models from large datasets with high accuracy.
method Introducing Isolation Kernel, which creates an exact, sparse, and finite-dimensional feature map of a kernel, allowing for efficient large scale online kernel learning without accuracy loss.
result Large scale online kernel learning can be achieved efficiently and accurately using Isolation Kernel.
Estimates kernel eigenvalues for compositional dot-product kernels.
problem Improving estimates for kernel eigenvalues.
method Eigenvalue decay estimates of integral operators associated with dot-product kernels.
result Improved estimates for kernel volumes in reproducing kernel Hilbert spaces.
Optimal Biweight kernel and computationally efficient Epanechnikov kernel for modal linear regression.
problem Finding the best kernel for modal linear regression.
method Refined analysis of asymptotic statistical behavior and IRLS algorithm convergence.
result Biweight kernel minimizes asymptotic mean squared error, Epanechnikov kernel guarantees IRLS convergence.
We present Random Partition Kernels, a new class of kernels derived by demonstrating a natural connection between random partitions of objects and kernels between those objects. We show how the construction can be used to create kernels from methods that would not normally be viewed as random partitions, such as Random…
The NNGP kernel's predictions closely match those of the Matern kernel under certain conditions.
problem Comparing NNGP kernels to Matern kernels in practical applications.
method Demonstrated the necessity of normalization for NNGP kernels, explored numerical challenges, and compared predictions and performance.
result NNGP kernel predictions closely match Matern kernel predictions under specific circumstances.
In this paper, we compare 5 different nonlinear kernels: min-max, RBF, fRBF (folded RBF), acos, and acos-χ2, on a wide range of publicly available datasets. The proposed fRBF kernel performs very similarly to the RBF kernel. Both RBF and fRBF kernels require an important tuning parameter (γ). Interestingly, for a …
New kernels allow learning from non-separable data.
problem Learning from non-separable data.
method Introducing entangled kernels and a two-step algorithm.
result Efficient algorithm for learning entangled kernels.
Laplace kernel and Neural Tangent Kernels are shown to be nearly identical for normalized data.
problem Understanding the similarity between Laplace and Neural Tangent Kernels.
method Theoretical analysis and experiments on normalized data.
result Laplace kernel and Neural Tangent Kernels have nearly identical eigenfunctions and RKHS for normalized data.
Kernel smoothing on unknown manifolds with bounds and asymptotic normality.
problem Data on unknown manifolds without boundaries.
method Finite sample bounds and asymptotic normality for kernel smoothing and its derivatives.
result Established finite sample bounds and asymptotic normality for kernel smoothing.
In this paper we propose a family of tractable kernels that is dense in the family of bounded positive semi-definite functions (i.e. can approximate any bounded kernel with arbitrary precision). We start by discussing the case of stationary kernels, and propose a family of spectral kernels that extends existing approac…
The success of kernel-based learning methods depend on the choice of kernel. Recently, kernel learning methods have been proposed that use data to select the most appropriate kernel, usually by combining a set of base kernels. We introduce a new algorithm for kernel learning that combines a {\em continuous set of base …
Quantum kernel machines need to use more complex kernels to fully exploit their potential.
problem Current quantum kernels struggle with complex learning tasks due to limited degrees of freedom.
method Propose using operator-valued kernels and C∗-algebraic representations to enhance quantum kernels. result Quantum operator-valued kernels can reveal structural dependencies that scalar-valued kernels miss.
New kernels capture both local and non-local interactions efficiently.
problem Designing kernels that capture both local and non-local interactions while remaining computationally tractable.
method Spectral truncation kernels based on C∗-algebra. result Spectral truncation kernels induce interactions across the data function domain and reduce computational cost.
Efficiently searches through Gaussian process kernels using symbolic representation and Bayesian optimization.
problem Manual selection of kernels in Gaussian processes is complex and computationally expensive.
method Proposes a novel method using symbolic representation and Bayesian optimization to search through a structured kernel space.
result Empirically shows a computationally more efficient way of searching through a discrete kernel space.
Optimal kernel improves estimation accuracy in modal statistical methods.
problem Estimation accuracy of kernel-based modal statistical methods depends on the kernel used.
method The study theoretically shows an optimal kernel that minimizes asymptotic error criterion.
result An optimal kernel minimizes the error criterion when using an optimal bandwidth.
The paper ensures stability of kernel methods under slight changes in probability measure, regularization, and kernel.
problem Stability of kernel-based methods under perturbations of probability measure, regularization, and kernel.
method Conditions for stability are derived based on convex Lipschitz loss functions and smooth kernels.
result Conditions for stability are given under simultaneous changes in probability measure, regularization, and kernel.