研究园简易版 · 经典书籍 · 期刊文章 · 有原文 · 完整版(可多选筛选、搜索) · 首页
2023 年的期刊文章 · 第 38 页
共 1511 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2023 到 2023。
按期刊、按主题看 →
JMLR Networks & Graphs Optimization & Learning Theory · 发表于 2023
Manoj Kumar, Anurag Sharma, Sandeep Kumar
摘要
Graph coarsening is a widely used dimensionality reduction technique for approaching large-scale graph machine-learning problems. Given a large graph, graph coarsening aims to learn a smaller-tractable graph while preserving the properties of the originally given graph. Graph data consist of node features and graph matrix (e.g., adjacency and Laplacian). The existing graph coarsening methods ignore the node features and rely solely on a graph matrix to simplify graphs. In this paper, we introduce a novel optimization-based framework for graph dimensionality reduction. The proposed framework lies in the unification of graph learning and dimensionality reduction. It takes both the graph matrix and the node features as the input and learns the coarsen graph matrix and the coarsen feature matrix jointly while ensuring desired properties. The proposed optimization formulation is a multi-block non-convex optimization problem, which is solved efficiently by leveraging block majorization-minimization, $\log$ determinant, Dirichlet energy, and regularization frameworks. The proposed algorithms are provably convergent and practically amenable to numerous tasks. It is also established that the learned coarsened graph is $\epsilon\in(0,1)$ similar to the original graph. Extensive experiments elucidate the efficacy of the proposed framework for real-world applications.
JMLR Machine Learning · 发表于 2023
Wei-Fang Sun, Cheng-Kuang Lee, Simon See, Chun-Yi Lee
摘要
In fully cooperative multi-agent reinforcement learning (MARL) settings, environments are highly stochastic due to the partial observability of each agent and the continuously changing policies of other agents. To address the above issues, we proposed a unified framework, called DFAC, for integrating distributional RL with value function factorization methods. This framework generalizes expected value function factorization methods to enable the factorization of return distributions. To validate DFAC, we first demonstrate its ability to factorize the value functions of a simple matrix game with stochastic rewards. Then, we perform experiments on all Super Hard maps of the StarCraft Multi-Agent Challenge and six self-designed Ultra Hard maps, showing that DFAC is able to outperform a number of baselines.
JMLR Causal Inference Networks & Graphs · 发表于 2023
Ehsan Mokhtarian, Saber Salehkaleybar, AmirEmad Ghassami, Negar Kiyavash
摘要
We study experiment design for unique identification of the causal graph of a simple SCM, where the graph may contain cycles. The presence of cycles in the structure introduces major challenges for experiment design as, unlike acyclic graphs, learning the skeleton of causal graphs with cycles may not be possible from merely the observational distribution. Furthermore, intervening on a variable in such graphs does not necessarily lead to orienting all the edges incident to it. In this paper, we propose an experiment design approach that can learn both cyclic and acyclic graphs and hence, unifies the task of experiment design for both types of graphs. We provide a lower bound on the number of experiments required to guarantee the unique identification of the causal graph in the worst case, showing that the proposed approach is order-optimal in terms of the number of experiments up to an additive logarithmic term. Moreover, we extend our result to the setting where the size of each experiment is bounded by a constant. For this case, we show that our approach is optimal in terms of the size of the largest experiment required for uniquely identifying the causal graph in the worst case.
JMLR Other · 发表于 2023
Haoyuan Sun, Khashayar Gatmiry, Kwangjun Ahn, Navid Azizan
摘要
Inspired by the remarkable success of large neural networks, there has been significant interest in understanding the generalization performance of over-parameterized models. Substantial efforts have been invested in characterizing how optimization algorithms impact generalization through their "preferred" solutions, a phenomenon commonly referred to as implicit regularization. In particular, it has been argued that gradient descent (GD) induces an implicit $\ell_2$-norm regularization in regression and classification problems. However, the implicit regularization of different algorithms are confined to either a specific geometry or a particular class of learning problems, indicating a gap in a general approach for controlling the implicit regularization. To address this, we present a unified approach using mirror descent (MD), a notable generalization of GD, to control implicit regularization in both regression and classification settings. More specifically, we show that MD with the general class of homogeneous potential functions converges in direction to a generalized maximum-margin solution for linear classification problems, thereby answering a long-standing question in the classification setting. Further, we show that MD can be implemented efficiently and enjoys fast convergence under suitable conditions. Through comprehensive experiments, we demonstrate that MD is a versatile method to produce learned models with different regularizers, which in turn have different generalization performances.
JMLR Other · 发表于 2023
Shiyuan He, Hanxuan Ye, Kejun He
摘要
This work studies the multi-task functional linear regression models where both the covariates and the unknown regression coefficients (called slope functions) are curves. For slope function estimation, we employ penalized splines to balance bias, variance, and computational complexity. The power of multi-task learning is brought in by imposing additional structures over the slope functions. We propose a general model with double regularization over the spline coefficient matrix: i) a matrix manifold constraint, and ii) a composite penalty as a summation of quadratic terms. Many multi-task learning approaches can be treated as special cases of this proposed model, such as a reduced-rank model and a graph Laplacian regularized model. We show the composite penalty induces a specific norm, which helps quantify the manifold curvature and determine the corresponding proper subset in the manifold tangent space. The complexity of tangent space subset is then bridged to the complexity of geodesic neighbor via generic chaining. A unified upper bound of the convergence rate is obtained and specifically applied to the reduced-rank model and the graph Laplacian regularized model. The phase transition behaviors for the estimators are examined as we vary the configurations of model parameters.
Econometrica Econometrics & Economics · 发表于 2023
Marek Pycia, Peter Troyan
摘要
We study extensive‐form games and mechanisms allowing agents that plan for only a subset of future decisions they may be called to make (the planning horizon ). Agents may update their so‐called strategic plan as the game progresses and new decision points enter their planning horizon. We introduce a family of simplicity standards which require that the prescribed action leads to unambiguously better outcomes, no matter what happens outside the planning horizon. We employ these standards to explore the trade‐off between simplicity and other objectives, to characterize simple mechanisms in a wide range of economic environments, and to delineate the simplicity of common mechanisms such as posted prices and ascending auctions, with the former being simpler than the latter.
JMLR Machine Learning Networks & Graphs · 发表于 2023
Jeremiah Zhe Liu, Shreyas Padhy, Jie Ren, Zi Lin, Yeming Wen, Ghassen Jerfel, Zachary Nado, Jasper Snoek, Dustin Tran, Balaji Lakshminarayanan
摘要
Accurate uncertainty quantification is a major challenge in deep learning, as neural networks can make overconfident errors and assign high confidence predictions to out-of-distribution (OOD) inputs. The most popular approaches to estimate predictive uncertainty in deep learning are methods that combine predictions from multiple neural networks, such as Bayesian neural networks (BNNs) and deep ensembles. However their practicality in real-time, industrial-scale applications are limited due to the high memory and computational cost. Furthermore, ensembles and BNNs do not necessarily fix all the issues with the underlying member networks. In this work, we study principled approaches to improve the uncertainty property of a single network, based on a single, deterministic representation. By formalizing the uncertainty quantification as a minimax learning problem, we first identify distance awareness, i.e., the model's ability to quantify the distance of a testing example from the training data, as a necessary condition for a DNN to achieve high-quality (i.e., minimax optimal) uncertainty estimation. We then propose Spectral-normalized Neural Gaussian Process (SNGP), a simple method that improves the distance-awareness ability of modern DNNs with two simple changes: (1) applying spectral normalization to hidden weights to enforce bi-Lipschitz smoothness in representations and (2) replacing the last output layer with a Gaussian process layer. On a suite of vision and language understanding benchmarks and on modern architectures (Wide-ResNet and BERT), SNGP consistently outperforms other single-model approaches in prediction, calibration and out-of-domain detection. Furthermore, SNGP provides complementary benefits to popular techniques such as deep ensembles and data augmentation, making it a simple and scalable building block for probabilistic deep learning.
Econometrica Other · 发表于 2023
Jean-Jacques Forneron
摘要
This paper proposes a Sieve Simulated Method of Moments (Sieve‐SMM) estimator for the parameters and the distribution of the shocks in nonlinear dynamic models where the likelihood and the moments are not tractable. An important concern with SMM, which matches sample with simulated moments, is that a parametric distribution is required. However, economic quantities that depend on this distribution, such as welfare and asset prices, can be sensitive to misspecification. The Sieve‐SMM estimator addresses this issue by flexibly approximating the distribution of the shocks with a Gaussian and tails mixture sieve. The asymptotic framework provides consistency, rate of convergence, and asymptotic normality results, extending existing results to a new framework with more general dynamics and latent variables. An application to asset pricing in a production economy shows a large decline in the estimates of relative risk aversion, highlighting the empirical relevance of misspecification bias.
JMLR Machine Learning · 发表于 2023
Nahuel Statuto, Irene Unceta, Jordi Nin, Oriol Pujol
摘要
Differential replication through copying refers to the process of replicating the decision behavior of a machine learning model using another model that possesses enhanced features and attributes. This process is relevant when external constraints limit the performance of an industrial predictive system. Under such circumstances, copying enables the retention of original prediction capabilities while adapting to new demands. Previous research has focused on the single-pass implementation for copying. This paper introduces a novel sequential approach that significantly reduces the amount of computational resources needed to train or maintain a copy, leading to reduced maintenance costs for companies using machine learning models in production. The effectiveness of the sequential approach is demonstrated through experiments with synthetic and real-world datasets, showing significant reductions in time and resources, while maintaining or improving accuracy.
JMLR High-Dimensional · 发表于 2023
Patricia Wollstadt, Sebastian Schmitt, Michael Wibral
摘要
Selecting a minimal feature set that is maximally informative about a target variable is a central task in machine learning and statistics. Information theory provides a powerful framework for formulating feature selection algorithms—yet, a rigorous, information-theoretic definition of feature relevancy, which accounts for feature interactions such as redundant and synergistic contributions, is still missing. We argue that this lack is inherent to classical information theory which does not provide measures to decompose the information a set of variables provides about a target into unique, redundant, and synergistic contributions. Such a decomposition has been introduced only recently by the partial information decomposition (PID) framework. Using PID, we clarify why feature selection is a conceptually difficult problem when approached using information theory and provide a novel definition of feature relevancy and redundancy in PID terms. From this definition, we show that the conditional mutual information (CMI) maximizes relevancy while minimizing redundancy and propose an iterative, CMI-based algorithm for practical feature selection. We demonstrate the power of our CMI-based algorithm in comparison to the unconditional mutual information on benchmark examples and provide corresponding PID estimates to highlight how PID allows to quantify information contribution of features and their interactions in feature-selection problems.
JMLR Other · 发表于 2023
Radu I. Bot, Michael Sedlmayer, Phan Tu Vuong
摘要
We introduce a relaxed inertial forward-backward-forward (RIFBF) splitting algorithm for approaching the set of zeros of the sum of a maximally monotone operator and a single-valued monotone and Lipschitz continuous operator. This work aims to extend Tseng's forward-backward-forward method by both using inertial effects as well as relaxation parameters. We formulate first a second order dynamical system that approaches the solution set of the monotone inclusion problem to be solved and provide an asymptotic analysis for its trajectories. We provide for RIFBF, which follows by explicit time discretization, a convergence analysis in the general monotone case as well as when applied to the solving of pseudo-monotone variational inequalities. We illustrate the proposed method by applications to a bilinear saddle point problem, in the context of which we also emphasize the interplay between the inertial and the relaxation parameters, and to the training of Generative Adversarial Networks (GANs).
JMLR High-Dimensional · 发表于 2023
Di Bo, Hoon Hwangbo, Vinit Sharma, Corey Arndt, Stephanie TerMaath
摘要
An analysis of high-dimensional data can offer a detailed description of a system but is often challenged by the curse of dimensionality. General dimensionality reduction techniques can alleviate such difficulty by extracting a few important features, but they are limited due to the lack of interpretability and connectivity to actual decision making associated with each physical variable. Variable selection techniques, as an alternative, can maintain the interpretability, but they often involve a greedy search that is susceptible to failure in capturing important interactions or a metaheuristic search that requires extensive computations. This research proposes a novel method that identifies critical subspaces, reduced-dimensional physical spaces, to achieve dimensionality reduction and variable selection. We apply a randomized search for subspace exploration and leverage ensemble techniques to enhance model performance. When applied to high-dimensional data collected from the failure prediction of a composite/metal hybrid structure exhibiting complex progressive damage failure under loading, the proposed method outperforms the existing and potential alternatives in prediction and important variable selection.
Econometrica Econometrics & Economics · 发表于 2023
Satyajit Chatterjee, Dean Corbae, Kyle Dempsey, José-Víctor Ríos-Rull
摘要
What is the role of credit scores in credit markets? We argue that it is, in part, the market's assessment of a person's unobservable type, which here we take to be patience. We postulate a model of persistent hidden types where observable actions shape the public assessment of a person's type via Bayesian updating. We show how dynamic reputation can incentivize repayment. Importantly, we show how an economy with credit scores implements the same equilibrium allocation. We estimate the model using both credit market data and the evolution of individuals' credit scores. We conduct counterfactuals to assess how more or less information used in scoring individuals affects outcomes and welfare. If tracking of individual credit actions is outlawed, poor young adults of low type benefit from subsidization by high types despite facing higher interest rates arising from lower dynamic incentives to repay.
JMLR Nonparametric & Semiparametric · 发表于 2023
Shubhanshu Shekhar, Ilmun Kim, Aaditya Ramdas
摘要
In nonparametric independence testing, we observe i.i.d.\ data $\{(X_i,Y_i)\}_{i=1}^n$, where $X \in \mathcal{X}, Y \in \mathcal{Y}$ lie in any general spaces, and we wish to test the null that $X$ is independent of $Y$. Modern test statistics such as the kernel Hilbert--Schmidt Independence Criterion (HSIC) and Distance Covariance (dCov) have intractable null distributions due to the degeneracy of the underlying U-statistics. Hence, in practice, one often resorts to using permutation testing, which provides a nonasymptotic guarantee at the expense of recalculating the quadratic-time statistics (say) a few hundred times. In this paper, we provide a simple but nontrivial modification of HSIC and dCov (called xHSIC and xdCov, pronounced “cross” HSIC/dCov) so that they have a limiting Gaussian distribution under the null, and thus do not require permutations. We show that our new tests, like the originals, are consistent against fixed alternatives, and minimax rate optimal against smooth local alternatives. Numerical simulations demonstrate that compared to the permutation tests, our variants have the same power within a constant factor, giving practitioners a new option for large problems or data-analysis pipelines where computation, not sample size, could be the bottleneck.
JMLR Other · 发表于 2023
Masaru Ito, Zhaosong Lu, Chuan He
摘要
In this paper we consider a composite optimization problem that minimizes the sum of a weakly smooth function and a convex function with either a bounded domain or a uniformly convex structure. In particular, we first present a parameter-dependent conditional gradient method for this problem, whose step sizes require prior knowledge of the parameters associated with the Hölder continuity of the gradient of the weakly smooth function, and establish its rate of convergence. Given that these parameters could be unknown or known but possibly conservative, such a method may suffer from implementation issue or slow convergence. We therefore propose a parameter-free conditional gradient method whose step size is determined by using a constructive local quadratic upper approximation and an adaptive line search scheme, without using any problem parameter. We show that this method achieves the same rate of convergence as the parameter-dependent conditional gradient method. Preliminary experiments are also conducted and illustrate the superior performance of the parameter-free conditional gradient method over the methods with some other step size rules.
JMLR Optimization & Learning Theory · 发表于 2023
Erhan Bayraktar, Ibrahim Ekren, Xin Zhang
摘要
In this paper, we study a learning problem in which a forecaster only observes partial information. By properly rescaling the problem, we heuristically derive a limiting PDE on Wasserstein space which characterizes the asymptotic behavior of the regret of the forecaster. Using a verification type argument, we show that the problem of obtaining regret bounds and efficient algorithms can be tackled by finding appropriate smooth sub/supersolutions of this parabolic PDE.
JMLR Other · 发表于 2023
Diego Delle Donne, Matthieu Kowalski, Leo Liberti
摘要
Given a vector $y \in \mathbb{R}^n$ and a matrix $H \in \mathbb{R}^{n\times m}$, the sparse approximation problem $\mathcal P_{0/p}$ asks for a point $x$ such that $\|y - Hx\|_p \leq \alpha$, for a given scalar $\alpha$, minimizing the size of the support $\|x\|_0 := \#\{j \ |\ x_j \neq 0 \}$. Existing convex mixed-integer programming formulations for $\mathcal P_{0/p}$ are of a kind referred to as “big-$M$”, meaning that they involve the use of a bound $M$ on the values of $x$. When a proper value for $M$ is not known beforehand, these formulations are not exact, in the sense that they may fail to recover the wanted global minimizer. In this work, we study the polytopes arising from these formulations and derive valid inequalities for them. We first use these inequalities to design a branch-and-cut algorithm for these models. Additionally, we prove that these inequalities are sufficient to describe the set of feasible supports for $\mathcal P_{0/p}$. Based on this result, we introduce a new (and the first to our knowledge) $M$-independent integer linear programming formulation for $\mathcal P_{0/p}$, which guarantees the recovery of the global minimizer. We propose a practical approach to tackle this formulation, which has exponentially many constraints. The proposed methods are then compared in computational experimentation to test their potential practical contribution.
JMLR Other · 发表于 2023
Lili Su, Jiaming Xu, Pengkun Yang
摘要
Federated Learning (FL) is a promising decentralized learning framework and has great potentials in privacy preservation and in lowering the computation load at the cloud. Recent work showed that FedAvg and FedProx -- the two widely-adopted FL algorithms -- fail to reach the stationary points of the global optimization objective even for homogeneous linear regression problems. Further, it is concerned that the common model learned might not generalize well locally at all in the presence of heterogeneity. In this paper, we analyze the convergence and statistical efficiency of FedAvg and FedProx, addressing the above two concerns. Our analysis is based on the standard non-parametric regression in a reproducing kernel Hilbert space (RKHS), and allows for heterogeneous local data distributions and unbalanced local datasets. We prove that the estimation errors, measured in either the empirical norm or the RKHS norm, decay with a rate of $1/t$ in general and exponentially for finite-rank kernels. In certain heterogeneous settings, these upper bounds also imply that both FedAvg and FedProx achieve the optimal error rate. To further analytically quantify the impact of the heterogeneity at each client, we propose and characterize a novel notion-federation gain, defined as the reduction of the estimation error for a client to join the FL. We discover that when the data heterogeneity is moderate, a client with limited local data can benefit from a common model with a large federation gain. Two new insights introduced by considering the statistical aspect are: (1) requiring the standard bounded dissimilarity is pessimistic for the convergence analysis of FedAvg and FedProx; (2) despite inconsistency of stationary points, their limiting points are unbiased estimators of the underlying truth. Numerical experiments further corroborate our theoretical findings.
JMLR Optimization & Learning Theory · 发表于 2023
Yasin Abbasi-Yadkori, András György, Nevena Lazić
摘要
We study the non-stationary stochastic multi-armed bandit problem, where the reward statistics of each arm may change several times during the course of learning. The performance of a learning algorithm is evaluated in terms of its dynamic regret, which is defined as the difference between the expected cumulative reward of an agent choosing the optimal arm in every time step and the cumulative reward of the learning algorithm. One way to measure the hardness of such environments is to consider how many times the identity of the optimal arm can change. We propose a method that achieves, in $K$-armed bandit problems, a near-optimal $\widetilde O(\sqrt{K N(S+1)})$ dynamic regret, where $N$ is the time horizon of the problem and $S$ is the number of times the identity of the optimal arm changes, without prior knowledge of $S$. Previous works for this problem obtain regret bounds that scale with the number of changes (or the amount of change) in the reward functions, which can be much larger, or assume prior knowledge of $S$ to achieve similar bounds.
JMLR Bayesian Methods · 发表于 2023
Michael J. O'Neill, Stephen J. Wright
摘要
We describe a line-search algorithm which achieves the best-known worst-case complexity results for problems with a certain “strict saddle” property that has been observed to hold in low-rank matrix optimization problems. Our algorithm is adaptive, in the sense that it makes use of backtracking line searches and does not require prior knowledge of the parameters that define the strict saddle property.
JMLR Machine Learning Nonparametric & Semiparametric · 发表于 2023
Minwoo Chae, Dongha Kim, Yongdai Kim, Lizhen Lin
摘要
We investigate statistical properties of a likelihood approach to nonparametric estimation of a singular distribution using deep generative models. More specifically, a deep generative model is used to model high-dimensional data that are assumed to concentrate around some low-dimensional structure. Estimating the distribution supported on this low-dimensional structure, such as a low-dimensional manifold, is challenging due to its singularity with respect to the Lebesgue measure in the ambient space. In the considered model, a usual likelihood approach can fail to estimate the target distribution consistently due to the singularity. We prove that a novel and effective solution exists by perturbing the data with an instance noise, which leads to consistent estimation of the underlying distribution with desirable convergence rates. We also characterize the class of distributions that can be efficiently estimated via deep generative models. This class is sufficiently general to contain various structured distributions such as product distributions, classically smooth distributions and distributions supported on a low-dimensional manifold. Our analysis provides some insights on how deep generative models can avoid the curse of dimensionality for nonparametric distribution estimation. We conduct a thorough simulation study and real data analysis to empirically demonstrate that the proposed data perturbation technique improves the estimation performance significantly.
JMLR Latent Variables & Mixtures · 发表于 2023
Haizi Yu, Igor Mineyev, Lav R. Varshney
摘要
Humans' abstraction ability plays a key role in concept learning and knowledge discovery. This theory paper presents the mathematical formulation for computationally emulating human-like abstractions---computational abstraction---and abstraction processes developed hierarchically from innate priors like symmetries. We study the nature of abstraction via a group-theoretic approach, formalizing and practically computing abstractions as symmetry-driven hierarchical clustering. Compared to data-driven clustering like k-means or agglomerative clustering (a chain), our abstraction model is data-free, feature-free, similarity-free, and globally hierarchical (a lattice). This paper also serves as a theoretical generalization of several existing works. These include generalizing Shannon's information lattice, specialized algorithms for certain symmetry-induced clusterings, as well as formalizing knowledge discovery applications such as learning music theory from scores and chemistry laws from molecules. We consider computational abstraction as a first step towards a principled and cognitive way of achieving human-level concept learning and knowledge discovery.
JMLR Meta-Analysis & Data Integration Optimization & Learning Theory · 发表于 2023
Yann Fraboni, Richard Vidal, Laetitia Kameni, Marco Lorenzi
摘要
We propose a novel framework to study asynchronous federated learning optimization with delays in gradient updates. Our theoretical framework extends the standard FedAvg aggregation scheme by introducing stochastic aggregation weights to represent the variability of the clients update time, due for example to heterogeneous hardware capabilities. Our formalism applies to the general federated setting where clients have heterogeneous datasets and perform at least one step of stochastic gradient descent (SGD). We demonstrate convergence for such a scheme and provide sufficient conditions for the related minimum to be the optimum of the federated problem. We show that our general framework applies to existing optimization schemes including centralized learning, FedAvg, asynchronous FedAvg, and FedBuff. The theory here provided allows drawing meaningful guidelines for designing a federated learning experiment in heterogeneous conditions. In particular, we develop in this work FedFix, a novel extension of FedAvg enabling efficient asynchronous federated training while preserving the convergence stability of synchronous aggregation. We empirically demonstrate our theory on a series of experiments showing that asynchronous FedAvg leads to fast convergence at the expense of stability, and we finally demonstrate the improvements of FedFix over synchronous and asynchronous FedAvg.
JMLR Causal Inference Networks & Graphs · 发表于 2023
Arrasy Rahman, Ignacio Carlucho, Niklas Höpner, Stefano V. Albrecht
摘要
Open ad hoc teamwork is the problem of training a single agent to efficiently collaborate with an unknown group of teammates whose composition may change over time. A variable team composition creates challenges for the agent, such as the requirement to adapt to new team dynamics and dealing with changing state vector sizes. These challenges are aggravated in real-world applications in which the controlled agent only has a partial view of the environment. In this work, we develop a class of solutions for open ad hoc teamwork under full and partial observability. We start by developing a solution for the fully observable case that leverages graph neural network architectures to obtain an optimal policy based on reinforcement learning. We then extend this solution to partially observable scenarios by proposing different methodologies that maintain belief estimates over the latent environment states and team composition. These belief estimates are combined with our solution for the fully observable case to compute an agent's optimal policy under partial observability in open ad hoc teamwork. Empirical results demonstrate that our solution can learn efficient policies in open ad hoc teamwork in fully and partially observable cases. Further analysis demonstrates that our methods' success is a result of effectively learning the effects of teammates' actions while also inferring the inherent state of the environment under partial observability.
JMLR Nonparametric & Semiparametric Networks & Graphs · 发表于 2023
David Holzmüller, Viktor Zaverkin, Johannes Kästner, Ingo Steinwart
摘要
The acquisition of labels for supervised learning can be expensive. To improve the sample efficiency of neural network regression, we study active learning methods that adaptively select batches of unlabeled data for labeling. We present a framework for constructing such methods out of (network-dependent) base kernels, kernel transformations, and selection methods. Our framework encompasses many existing Bayesian methods based on Gaussian process approximations of neural networks as well as non-Bayesian methods. Additionally, we propose to replace the commonly used last-layer features with sketched finite-width neural tangent kernels and to combine them with a novel clustering method. To evaluate different methods, we introduce an open-source benchmark consisting of 15 large tabular regression data sets. Our proposed method outperforms the state-of-the-art on our benchmark, scales to large data sets, and works out-of-the-box without adjusting the network architecture or training code. We provide open-source code that includes efficient implementations of all kernels, kernel transformations, and selection methods, and can be used for reproducing our results.
JMLR Optimization & Learning Theory · 发表于 2023
Kexin Jin, Jonas Latz, Chenguang Liu, Carola-Bibiane Schönlieb
摘要
Optimization problems with continuous data appear in, e.g., robust machine learning, functional data analysis, and variational inference. Here, the target function is given as an integral over a family of (continuously) indexed target functions---integrated with respect to a probability measure. Such problems can often be solved by stochastic optimization methods: performing optimization steps with respect to the indexed target function with randomly switched indices. In this work, we study a continuous-time variant of the stochastic gradient descent algorithm for optimization problems with continuous data. This so-called stochastic gradient process consists in a gradient flow minimizing an indexed target function that is coupled with a continuous-time index process determining the index. Index processes are, e.g., reflected diffusions, pure jump processes, or other Lévy processes on compact spaces. Thus, we study multiple sampling patterns for the continuous data space and allow for data simulated or streamed at runtime of the algorithm. We analyze the approximation properties of the stochastic gradient process and study its longtime behavior and ergodicity under constant and decreasing learning rates. We end with illustrating the applicability of the stochastic gradient process in a polynomial regression problem with noisy functional data, as well as in a physics-informed neural network.
JMLR Other · 发表于 2023
Juan C. Perdomo, Akshay Krishnamurthy, Peter Bartlett, Sham Kakade
摘要
Offline policy evaluation is a fundamental statistical problem in reinforcement learning that involves estimating the value function of some decision-making policy given data collected by a potentially different policy. In order to tackle problems with complex, high-dimensional observations, there has been significant interest from theoreticians and practitioners alike in understanding the possibility of function approximation in reinforcement learning. Despite significant study, a sharp characterization of when we might expect offline policy evaluation to be tractable, even in the simplest setting of linear function approximation, has so far remained elusive, with a surprising number of strong negative results recently appearing in the literature. In this work, we identify simple control-theoretic and linear-algebraic conditions that are necessary and sufficient for classical methods, in particular Fitted Q-iteration (FQI) and least squares temporal difference learning (LSTD), to succeed at offline policy evaluation. Using this characterization, we establish a precise hierarchy of regimes under which these estimators succeed. We prove that LSTD works under strictly weaker conditions than FQI. Furthermore, we establish that if a problem is not solvable via LSTD, then it cannot be solved by a broad class of linear estimators, even in the limit of infinite data. Taken together, our results provide a complete picture of the behavior of linear estimators for offline policy evaluation, unify previously disparate analyses of canonical algorithms, and provide significantly sharper notions of the underlying statistical complexity of offline policy evaluation.
Econometrica Econometrics & Economics · 发表于 2023
Craig A. Chikis, Jonathan Goldberg, David López-Salido
摘要
Using an endogenous growth model, Liu, Mian, and Sufi (2022) (LMS) show that a decline in the interest rate can lead to a fall in productivity growth and a rise in leader‐laggard productivity gaps and firm profits. We identify two issues in their quantitative analysis of transition dynamics: a time‐scale error and the omission of composition terms in calculating productivity growth along the transition to a new balanced growth path. Correcting the time‐scale error and including the composition terms, the decline in the interest rate that LMS study leads to a large and protracted productivity boom lasting about 20 years. In addition, the average leader‐laggard gap grows much more slowly than reported in their paper. We also point out an issue in their quantitative analysis of steady‐state profit shares. These issues are related to the quantitative exercises, and do not affect the key theoretical contributions of LMS.
Econometrica Other · 发表于 2023
Keisuke Hirano
摘要
This paper discussed “Invidious Comparisons: Ranking and Selection as Compound Decisions” by Jiaying Gu and Roger Koenker. *(abstract via OpenAlex)*
JMLR Bayesian Methods · 发表于 2023
Jacques Wainer
摘要
his paper presents a Bayesian model, called the Bayesian Bradley Terry (BBT) model, for comparing multiple algorithms on multiple data sets based on any metric. The model is an extension of the Bradley Terry model, which tracks the number of wins each algorithm has on different data sets. Unlike frequentist methods such as Demsar tests on mean rank or multiple pairwise Wilcoxon tests, the Bayesian approach provides a more nuanced understanding of the algorithms’ performance and allows for the definition of the “region of practical equivalence” (ROPE) for two algorithms. Additionally, the paper introduces the concept of “local ROPE,” which assesses the significance of the difference in mean measure between two algorithms using effect sizes, and can be applied in frequentist approaches as well. Both an R package and a Python program implementing the BBT are available for use.