2017 年的期刊文章 · 第 27 页

共 1233 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2017 到 2017。

按期刊、按主题看 →

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 页

Econometrica Other · 发表于 2017

The Identification of Beliefs From Asset Demand

Felix Kübler, Herakles Polemarchakis

摘要

The demand for assets as prices and initial wealth vary identifies beliefs and attitudes towards risk.We derive conditions that guarantee identification with no knowledge either of the cardinal utility index (attitudes towards risk) or of the distribution of future endowments or payoffs of assets; the argument applies even if the asset market is incomplete and demand is observed only locally. *(abstract via OpenAlex)*

Econometrica Other · 发表于 2017

The Evolution of Culture and Institutions: Evidence From the Kuba Kingdom

Sara Lowes, Nathan Nunn, James A. Robinson, Jonathan L. Weigel

摘要

We use variation in historical state centralization to examine the long-term impact of institutions on cultural norms. The Kuba Kingdom, established in Central Africa in the early 17th century by King Shyaam, had more developed state institutions than the other independent villages and chieftaincies in the region. It had an unwritten constitution, separation of political powers, a judicial system with courts and juries, a police force, a military, taxation, and significant public goods provision. Comparing individuals from the Kuba Kingdom to those from just outside the Kingdom, we find that centralized formal institutions are associated with weaker norms of rule following and a greater propensity to cheat for material gain. This finding is consistent with recent models where endogenous investments to inculcate values in children decline when there is an increase in the effectiveness of formal institutions that enforce socially desirable behavior. Consistent with such a mechanism, we find that Kuba parents believe it is less important to teach children values related to rule-following behaviors. *(abstract via OpenAlex)*

JMLR Other · 发表于 2017

Tests of Mutual or Serial Independence of Random Vectors with Applications

Martin Bilodeau, Aurélien Guetsop Nangue

摘要

The problem of testing mutual independence between many random vectors is addressed. The closely related problem of testing serial independence of a multivariate stationary sequence is also considered. The Möbius transformation of characteristic functions is used to characterize independence. A generalization to $p$ vectors of distance covariance and Hilbert-Schmidt independence criterion ($HSIC$) tests with the translation invariant kernel of a stable probability distribution is proposed. Both test statistics can be expressed in a simple form as a sum over all elements of a componentwise product of $p$ doubly-centered matrices. It is shown that an $HSIC$ statistic with sufficiently small scale parameters is equivalent to a distance covariance statistic. Consistency and weak convergence of both types of statistics are established. Approximation of $p$-values is made by randomization tests without recomputing interpoint distances for each randomized sample. The dependogram is adapted to the proposed tests for the graphical identification of sources of dependencies. Empirical rejection rates obtained through extensive simulations confirm both the applicability of the testing procedures in small samples and the high level of competitiveness in terms of power. Applications to meteorological and financial data provide some interesting interpretations of dependencies revealed by dependograms.

JMLR Networks & Graphs · 发表于 2017

Target Curricula via Selection of Minimum Feature Sets: a Case Study in Boolean Networks

Shannon Fenn, Pablo Moscato

摘要

We consider the effect of introducing a curriculum of targets when training Boolean models on supervised Multi Label Classification (MLC) problems. In particular, we consider how to order targets in the absence of prior knowledge, and how such a curriculum may be enforced when using meta-heuristics to train discrete non-linear models.

Econometrica Other · 发表于 2017

Strong Duality for a Multiple-Good Monopolist

Constantinos Daskalakis, Alan Deckelbaum, Christos Tzamos

摘要

We characterize optimal mechanisms for the multiple-good monopoly problem and provide a framework to find them. We show that a mechanism is optimal if and only if a measure $μ$ derived from the buyer's type distribution satisfies certain stochastic dominance conditions. This measure expresses the marginal change in the seller's revenue under marginal changes in the rent paid to subsets of buyer types. As a corollary, we characterize the optimality of grand-bundling mechanisms, strengthening several results in the literature, where only sufficient optimality conditions have been derived. As an application, we show that the optimal mechanism for $n$ independent uniform items each supported on $[c,c+1]$ is a grand-bundling mechanism, as long as $c$ is sufficiently large, extending Pavlov's result for $2$ items [Pavlov'11]. At the same time, our characterization also implies that, for all $c$ and for all sufficiently large $n$, the optimal mechanism for $n$ independent uniform items supported on $[c,c+1]$ is not a grand bundling mechanism. *(abstract via OpenAlex)*

JMLR Other · 发表于 2017

Stochastic Primal-Dual Coordinate Method for Regularized Empirical Risk Minimization

Yuchen Zhang, Lin Xiao

摘要

We consider a generic convex optimization problem associated with regularized empirical risk minimization of linear predictors. The problem structure allows us to reformulate it as a convex-concave saddle point problem. We propose a stochastic primal-dual coordinate (SPDC) method, which alternates between maximizing over a randomly chosen dual variable and minimizing over the primal variables. An extrapolation step on the primal variables is performed to obtain accelerated convergence rate. We also develop a mini-batch version of the SPDC method which facilitates parallel computing, and an extension with weighted sampling probabilities on the dual variables, which has a better complexity than uniform sampling on unnormalized data. Both theoretically and empirically, we show that the SPDC method has comparable or better performance than several state-of-the-art optimization methods.

JMLR Bayesian Methods Optimization & Learning Theory · 发表于 2017

Stochastic Gradient Descent as Approximate Bayesian Inference

Stephan Mandt, Matthew D. Hoffman, David M. Blei

摘要

Stochastic Gradient Descent with a constant learning rate (constant SGD) simulates a Markov chain with a stationary distribution. With this perspective, we derive several new results. (1) We show that constant SGD can be used as an approximate Bayesian posterior inference algorithm. Specifically, we show how to adjust the tuning parameters of constant SGD to best match the stationary distribution to a posterior, minimizing the Kullback-Leibler divergence between these two distributions. (2) We demonstrate that constant SGD gives rise to a new variational EM algorithm that optimizes hyperparameters in complex probabilistic models. (3) We also show how to tune SGD with momentum for approximate sampling. (4) We analyze stochastic-gradient MCMC algorithms. For Stochastic- Gradient Langevin Dynamics and Stochastic-Gradient Fisher Scoring, we quantify the approximation errors due to finite learning rates. Finally (5), we use the stochastic process perspective to give a short proof of why Polyak averaging is optimal. Based on this idea, we propose a scalable approximate MCMC algorithm, the Averaged Stochastic Gradient Sampler.

JMLR Latent Variables & Mixtures · 发表于 2017

Statistical and Computational Guarantees for the Baum-Welch Algorithm

Fanny Yang, Sivaraman Balakrishnan, Martin J. Wainwright

摘要

The Hidden Markov Model (HMM) is one of the mainstays of statistical modeling of discrete time series, with applications including speech recognition, computational biology, computer vision and econometrics. Estimating an HMM from its observation process is often addressed via the Baum-Welch algorithm, which is known to be susceptible to local optima. In this paper, we first give a general characterization of the basin of attraction associated with any global optimum of the population likelihood. By exploiting this characterization, we provide non-asymptotic finite sample guarantees on the Baum-Welch updates and show geometric convergence to a small ball of radius on the order of the minimax rate around a global optimum. As a concrete example, we prove a linear rate of convergence for a hidden Markov mixture of two isotropic Gaussians given a suitable mean separation and an initialization within a ball of large radius around (one of) the true parameters. To our knowledge, these are the first rigorous local convergence guarantees to global optima for the Baum-Welch algorithm in a setting where the likelihood function is nonconvex. We complement our theoretical results with thorough numerical simulations studying the convergence of the Baum-Welch algorithm and illustrating the accuracy of our predictions.

Econometrica Other · 发表于 2017

Statistical Properties of Microstructure Noise

Jean Jacod, Yingying Li, Xinghua Zheng

摘要

We study the estimation of (joint) moments of microstructure noise based on high frequency data. The estimation is conducted under a nonparametric setting, which allows the underlying price process to have jumps, the observation times to be irregularly spaced, and the noise to be dependent on the price process and to have diurnal features. Estimators of arbitrary orders of (joint) moments are provided, for which we establish consistency as well as central limit theorems. In particular, we provide estimators of autocovariances and autocorrelations of the noise. Simulation studies demonstrate excellent performance of our estimators in the presence of jumps, irregular observation times, and even rounding. Empirical studies reveal (moderate) positive autocorrelations of microstructure noise for the stocks tested. *(abstract via OpenAlex)*

JMLR Other · 发表于 2017

Statistical Inference with Unnormalized Discrete Models and Localized Homogeneous Divergences

Takashi Takenouchi, Takafumi Kanamori

摘要

In this paper, we focus on parameters estimation of probabilistic models in discrete space. A naive calculation of the normalization constant of the probabilistic model on discrete space is often infeasible and statistical inference based on such probabilistic models has difficulty. In this paper, we propose a novel estimator for probabilistic models on discrete space, which is derived from an empirically localized homogeneous divergence. The idea of the empirical localization makes it possible to ignore an unobserved domain on sample space, and the homogeneous divergence is a discrepancy measure between two positive measures and has a weak coincidence axiom. The proposed estimator can be constructed without calculating the normalization constant and is asymptotically consistent and Fisher efficient. We investigate statistical properties of the proposed estimator and reveal a relationship between the empirically localized homogeneous divergence and a mixture of the $\alpha$-divergence. The $\alpha$-divergence is a non- homogeneous discrepancy measure that is frequently discussed in the context of information geometry. Using the relationship, we also propose an asymptotically consistent estimator of the normalization constant. Experiments showed that the proposed estimator comparably performs to the maximum likelihood estimator but with drastically lower computational cost.

JMLR High-Dimensional Optimization & Learning Theory · 发表于 2017

Stabilized Sparse Online Learning for Sparse Data

Yuting Ma, Tian Zheng

摘要

Stochastic gradient descent (SGD) is commonly used for optimization in large-scale machine learning problems. Lanford et al. (2009) introduce a sparse online learning method to induce sparsity via truncated gradient. With high- dimensional sparse data, however, this method suffers from slow convergence and high variance due to heterogeneity in feature sparsity. To mitigate this issue, we introduce a stabilized truncated stochastic gradient descent algorithm. We employ a soft- thresholding scheme on the weight vector where the imposed shrinkage is adaptive to the amount of information available in each feature. The variability in the resulted sparse weight vector is further controlled by stability selection integrated with the informative truncation. To facilitate better convergence, we adopt an annealing strategy on the truncation rate, which leads to a balanced trade-off between exploration and exploitation in learning a sparse weight vector. Numerical experiments show that our algorithm compares favorably with the original truncated gradient SGD in terms of prediction accuracy, achieving both better sparsity and stability.

JMLR Other · 发表于 2017

Stability of Controllers for Gaussian Process Dynamics

Julia Vinogradska, Bastian Bischoff, Duy Nguyen-Tuong, Jan Peters

摘要

Learning control has become an appealing alternative to the derivation of control laws based on classic control theory. However, a major shortcoming of learning control is the lack of performance guarantees which prevents its application in many real-world scenarios. As a step towards widespread deployment of learning control, we provide stability analysis tools for controllers acting on dynamics represented by Gaussian processes (GPs). We consider differentiable Markovian control policies and system dynamics given as (i) the mean of a GP, and (ii) the full GP distribution. For both cases, we analyze finite and infinite time horizons. Furthermore, we study the effect of disturbances on the stability results. Empirical evaluations on simulated benchmark problems support our theoretical results.

Econometrica Other · 发表于 2017

Spurious Inference in Reduced-Rank Asset-Pricing Models

Nikolay Gospodinov, Raymond Kan, Cesare Robotti

摘要

We study some seemingly anomalous results that arise in possibly misspecified, reduced-rank linear asset-pricing models estimated by the continuously updated generalized method of moments. When a spurious factor (that is, a factor that is uncorrelated with the returns on the test assets) is present, the test for correct model specification has asymptotic power that is equal to the nominal size. In other words, applied researchers will erroneously conclude that the model is correctly specified even when the degree of misspecification is arbitrarily large.The rejection probability of the test for overidentifying restrictions typically decreases further in underidentified models where the dimension of the null space is larger than 1. *(abstract via OpenAlex)*

JMLR Latent Variables & Mixtures Networks & Graphs · 发表于 2017

Spectral Clustering Based on Local PCA

Ery Arias-Castro, Gilad Lerman, Teng Zhang

摘要

We propose a spectral clustering method based on local principal components analysis (PCA). After performing local PCA in selected neighborhoods, the algorithm builds a nearest neighbor graph weighted according to a discrepancy between the principal subspaces in the neighborhoods, and then applies spectral clustering. As opposed to standard spectral methods based solely on pairwise distances between points, our algorithm is able to resolve intersections. We establish theoretical guarantees for simpler variants within a prototypical mathematical framework for multi-manifold clustering, and evaluate our algorithm on various simulated data sets.

JMLR Optimization & Learning Theory · 发表于 2017

Soft Margin Support Vector Classification as Buffered Probability Minimization

Matthew Norton, Alexander Mafusalov, Stan Uryasev

摘要

In this paper, we show that the popular C-SVM, soft-margin support vector classifier is equivalent to minimization of Buffered Probability of Exceedance (bPOE), a recently introduced characterization of uncertainty. To show this, we introduce a new SVM formulation, called the EC-SVM, which is derived from a simple bPOE minimization problem that is easy to interpret with a meaningful free parameter, optimal objective value, and probabilistic derivation. Over the range of its free parameter, the EC-SVM has both a convex and non-convex case which we connect to existing SVM formulations. We first show that the C-SVM, formulated with any regularization norm, is equivalent to the convex EC-SVM. Similarly, we show that the E$\nu$-SVM is equivalent to the EC-SVM over its entire parameter range, which includes both the convex and non-convex case. These equivalences, coupled with the interpretability of the EC-SVM, allow us to gain surprising new insights into the C-SVM and fully connect soft margin support vector classification with superquantile and bPOE concepts. We also show that the EC-SVM can easily be cast as a robust optimization problem, where bPOE is minimized with data lying in a fixed uncertainty set. This reformulation allows us to clearly differentiate between the convex and non-convex case, with convexity associated with pessimistic views of uncertainty and non-convexity associated with optimistic views of uncertainty. Finally, we address some practical considerations. First, we show that these new insights can assist in making parameter selection more efficient. Second, we discuss optimization approaches for solving the EC-SVM. Third, we address the issue of generalization, providing generalization bounds for both bPOE and misclassification rate.

JMLR Optimization & Learning Theory Networks & Graphs · 发表于 2017

SnapVX: A Network-Based Convex Optimization Solver

David Hallac, Christopher Wong, Steven Diamond, Abhijit Sharang, Rok Sosič, Stephen Boyd, Jure Leskovec

摘要

SnapVX is a high-performance solver for convex optimization problems defined on networks. For problems of this form, SnapVX provides a fast and scalable solution with guaranteed global convergence. It combines the capabilities of two open source software packages: Snap.py and CVXPY. Snap.py is a large scale graph processing library, and CVXPY provides a general modeling framework for small-scale subproblems. SnapVX offers a customizable yet easy-to-use Python interface with out-of- the- box functionality. Based on the Alternating Direction Method of Multipliers (ADMM), it is able to efficiently store, analyze, parallelize, and solve large optimization problems from a variety of different applications. Documentation, examples, and more can be found on the SnapVX website at snap.stanford.edu/snapvx.

Econometrica Other · 发表于 2017

Single-Crossing Random Utility Models

Jose Apesteguia, Miguel A. Ballester, Jay Lu

摘要

We propose a novel model of stochastic choice: the single-crossing random utility model (SCRUM). This is a random utility model in which the collection of preferences satisfies the single-crossing property. We o↵er a characterization of SCRUMs based on two easy-to-check properties: the classic Monotonicity property and a novel condition, Centrality. The identified collection of preferences and associated probabilities is unique. We show that SCRUMs nest both single-peaked and single-dipped random utility models and establish a stochastic monotone comparative result for the case of SCRUMs. *(abstract via OpenAlex)*

JMLR Causal Inference · 发表于 2017

Simplifying Probabilistic Expressions in Causal Inference

Santtu Tikka, Juha Karvanen

摘要

Obtaining a non-parametric expression for an interventional distribution is one of the most fundamental tasks in causal inference. Such an expression can be obtained for an identifiable causal effect by an algorithm or by manual application of do-calculus. Often we are left with a complicated expression which can lead to biased or inefficient estimates when missing data or measurement errors are involved. We present an automatic simplification algorithm that seeks to eliminate symbolically unnecessary variables from these expressions by taking advantage of the structure of the underlying graphical model. Our method is applicable to all causal effect formulas and is readily available in the R package causaleffect.

JMLR High-Dimensional · 发表于 2017

Sharp Oracle Inequalities for Square Root Regularization

Benjamin Stucky, Sara van de Geer

摘要

We study a set of regularization methods for high-dimensional linear regression models. These penalized estimators have the square root of the residual sum of squared errors as loss function, and any weakly decomposable norm as penalty function. This fit measure is chosen because of its property that the estimator does not depend on the unknown standard deviation of the noise. On the other hand, a generalized weakly decomposable norm penalty is very useful in being able to deal with different underlying sparsity structures. We can choose a different sparsity inducing norm depending on how we want to interpret the unknown parameter vector $\beta$. Structured sparsity norms, as defined in Micchelli et al. (2010), are special cases of weakly decomposable norms, therefore we also include the square root LASSO (Belloni et al., 2011), the group square root LASSO (Bunea et al., 2014) and a new method called the square root SLOPE (in a similar fashion to the SLOPE from Bogdan et al. 2015). For this collection of estimators our results provide sharp oracle inequalities with the Karush-Kuhn-Tucker conditions. We discuss some examples of estimators. Based on a simulation we illustrate some advantages of the square root SLOPE.

JMLR Machine Learning Optimization & Learning Theory · 发表于 2017

Second-Order Stochastic Optimization for Machine Learning in Linear Time

Naman Agarwal, Brian Bullins, Elad Hazan

摘要

First-order stochastic methods are the state-of-the-art in large-scale machine learning optimization owing to efficient per-iteration complexity. Second-order methods, while able to provide faster convergence, have been much less explored due to the high cost of computing the second-order information. In this paper we develop second-order stochastic methods for optimization problems in machine learning that match the per- iteration cost of gradient based methods, and in certain settings improve upon the overall running time over popular first-order methods. Furthermore, our algorithm has the desirable property of being implementable in time linear in the sparsity of the input data.

Econometrica Other · 发表于 2017

Search for Yield

David Martinez-Miera, Rafael Repullo

摘要

We present a model of the relationship between real interest rates, credit spreads, and the structure and risk of the banking system. Banks intermediate between entrepreneurs and investors, and can monitor entrepreneurs projects. We characterize the equilibrium for a xed aggregate supply of savings, showing that safer entrepreneurs \nwill be funded by nonmonitoring banks and riskier entrepreneurs by monitoring banks. We show that an increase in savings reduces interest rates and spreads, increases the relative size of the nonmonitoring banking system and the probability of failure of monitoring banks. We also show that the dynamic version of the model exhibits endogenous boom and bust cycles, and rationalizes the existence of countercyclical risk premia and the connection between low interest rates, credit spreads, and the buildup of risks during booms. *(abstract via OpenAlex)*

JMLR Networks & Graphs · 发表于 2017

Scalable Influence Maximization for Multiple Products in Continuous-Time Diffusion Networks

Nan Du, Yingyu Liang, Maria-Florina Balcan, Manuel Gomez-Rodriguez, Hongyuan Zha, Le Song

摘要

A typical viral marketing model identifies influential users in a social network to maximize a single product adoption assuming unlimited user attention, campaign budgets, and time. In reality, multiple products need campaigns, users have limited attention, convincing users incurs costs, and advertisers have limited budgets and expect the adoptions to be maximized soon. Facing these user, monetary, and timing constraints, we formulate the problem as a submodular maximization task in a continuous-time diffusion model under the intersection of one matroid and multiple knapsack constraints. We propose a randomized algorithm estimating the user influence (Partial results in the paper on influence estimation have been published in a conference paper: Nan Du, Le Song, Manuel Gomez-Rodriguez, and Hongyuan Zha. Scalable influence estimation in continuous time diffusion networks. In Advances in Neural Information Processing Systems 26, 2013.) in a network ($|\mathcal{V}|$ nodes, $|\mathcal{E}|$ edges) to an accuracy of $\epsilon$ with $n=\mathcal{O}(1/\epsilon^2)$ randomizations and $\tilde{\mathcal{O}}(n|\mathcal{E}|+n|\mathcal{V}|)$ computations. By exploiting the influence estimation algorithm as a subroutine, we develop an adaptive threshold greedy algorithm achieving an approximation factor $k_a/(2+2 k)$ of the optimal when $k_a$ out of the $k$ knapsack constraints are active. Extensive experiments on networks of millions of nodes demonstrate that the proposed algorithms achieve the state-of- the-art in terms of effectiveness and scalability.

Econometrica Econometrics & Economics · 发表于 2017

Sales Force and Competition in Financial Product Markets: The Case of Mexico's Social Security Privatization

Justine Hastings, Ali Hortaçsu, Chad Syverson

摘要

This paper examines how sales force impacts competition and equilibrium prices in the context of a privatized pension market. We use detailed administrative data on fund manager choices and worker characteristics at the inception of Mexico's privatized social security system, where fund managers had to set prices (management fees) at the national level, but could select sales force levels by local geographic areas. We develop and estimate a model of fund manager choice where sales force can increase or decrease customer price sensitivity. We find exposure to sales force lowered price sensitivity, leading to inelastic demand and high equilibrium fees. We simulate oft proposed policy solutions: a supply‐side policy with a competitive government player and a demand‐side policy that increases price elasticity. We find that demand‐side policies are necessary to foster competition in social safety net markets with large segments of inelastic consumers. *(abstract via OpenAlex)*

JMLR High-Dimensional · 发表于 2017

STORE: Sparse Tensor Response Regression and Neuroimaging Analysis

Will Wei Sun, Lexin Li

摘要

Motivated by applications in neuroimaging analysis, we propose a new regression model, Sparse TensOr REsponse regression (STORE), with a tensor response and a vector predictor. STORE embeds two key sparse structures: element-wise sparsity and low-rankness. It can handle both a non-symmetric and a symmetric tensor response, and thus is applicable to both structural and functional neuroimaging data. We formulate the parameter estimation as a non-convex optimization problem, and develop an efficient alternating updating algorithm. We establish a non- asymptotic estimation error bound for the actual estimator obtained from the proposed algorithm. This error bound reveals an interesting interaction between the computational efficiency and the statistical rate of convergence. When the distribution of the error tensor is Gaussian, we further obtain a fast estimation error rate which allows the tensor dimension to grow exponentially with the sample size. We illustrate the efficacy of our model through intensive simulations and an analysis of the Autism spectrum disorder neuroimaging data.

Econometrica Econometrics & Economics · 发表于 2017

Rushes in Large Timing Games

Axel Anderson, Lones Smith, Andreas Park

摘要

We develop a continuum player timing game that subsumes standard wars of attrition and pre-emption games, and introduces a new rushes phenomenon. Payoffs are continuous and single-peaked functions of the stopping time and stopping quantile. We show that if payoffs are hump-shaped in the quantile, then a sudden “rush” of players stops in any Nash or subgame perfect equilibrium. Fear relaxes the first mover advantage in pre-emption games, asking that the least quantile beat the average; greed relaxes the last mover advantage in wars of attrition, asking just that the last quantile payoff exceed the average. With greed, play is inefficiently late: an accelerating war of attrition starting at optimal time, followed by a rush. With fear, play is inefficiently early: a slowing pre-emption game, ending at the optimal time, preceded by a rush. The theory predicts the length, duration, and intensity of stopping, and the size and timing of rushes, and offers insights for many common timing games. *(abstract via OpenAlex)*

Econometrica Other · 发表于 2017

Robustness and Separation in Multidimensional Screening

Gabriel Carroll

摘要

A principal wishes to screen an agent along several dimensions simultaneously. The agent has quasilinear preferences that are additively separable across the various components. We consider a robust version of the principal’s problem, in which she knows the marginal distribution of each component of the agent’s type, but does not know the joint distribution. Any mechanism is evaluated by its worst-case expected profit, over all joint distributions consistent with the known marginals. We show that the optimum for the principal is simply to screen along each component separately. This result does not require any assumptions (such as single-crossing) on the structure of preferences within each component. Applications of the model include monopoly pricing and dynamic taxation. This paper has greatly benefited from conversations with Florian Scheuer, as well as helpful comments from (in random order) Richard Holden, Dawen Meng, Andy *(abstract via OpenAlex)*

JMLR Bayesian Methods · 发表于 2017

Robust and Scalable Bayes via a Median of Subset Posterior Measures

Stanislav Minsker, Sanvesh Srivastava, Lizhen Lin, David B. Dunson

摘要

We propose a novel approach to Bayesian analysis that is provably robust to outliers in the data and often has computational advantages over standard methods. Our technique is based on splitting the data into non-overlapping subgroups, evaluating the posterior distribution given each independent subgroup, and then combining the resulting measures. The main novelty of our approach is the proposed aggregation step, which is based on the evaluation of a median in the space of probability measures equipped with a suitable collection of distances that can be quickly and efficiently evaluated in practice. We present both theoretical and numerical evidence illustrating the improvements achieved by our method.

JMLR High-Dimensional Latent Variables & Mixtures · 发表于 2017

Robust Discriminative Clustering with Sparse Regularizers

Nicolas Flammarion, Balamurugan Palaniappan, Francis Bach

摘要

Clustering high-dimensional data often requires some form of dimensionality reduction, where clustered variables are separated from noise-looking variables. We cast this problem as finding a low-dimensional projection of the data which is well-clustered. This yields a one-dimensional projection in the simplest situation with two clusters, and extends naturally to a multi-label scenario for more than two clusters. In this paper, (a) we first show that this joint clustering and dimension reduction formulation is equivalent to previously proposed discriminative clustering frameworks, thus leading to convex relaxations of the problem; (b) we propose a novel sparse extension, which is still cast as a convex relaxation and allows estimation in higher dimensions; (c) we propose a natural extension for the multi-label scenario; (d) we provide a new theoretical analysis of the performance of these formulations with a simple probabilistic model, leading to scalings over the form $d=O(\sqrt{n})$ for the affine invariant case and $d=O(n)$ for the sparse case, where $n$ is the number of examples and $d$ the ambient dimension; and finally, (e) we propose an efficient iterative algorithm with running-time complexity proportional to $O(nd^2)$, improving on earlier algorithms for discriminative clustering with the square loss, which had quadratic complexity in the number of examples.

Econometrica Causal Inference Testing & Uncertainty · 发表于 2017

Robust Confidence Intervals for Average Treatment Effects Under Limited Overlap

Christoph Rothe

摘要

Robust Confidence Intervals for Average Treatment Effects under Limited Overlap *Estimators of average treatment effects under unconfounded treatment assignment are known to become rather imprecise if there is limited overlap in the covariate distributions between the treatment groups.But such limited overlap can also have a detrimental effect on inference, and lead for example to highly distorted confidence intervals.This paper shows that this is because the coverage error of traditional confidence intervals is not so much driven by the total sample size, but by the number of observations in the areas of limited overlap.At least some of these "local sample sizes" are often very small in applications, up to the point where distributional approximation derived from the Central Limit Theorem become unreliable.Building on this observation, the paper proposes two new robust confidence intervals that are extensions of classical approaches to small sample inference.It shows that these approaches are easy to implement, and have superior theoretical and practical properties relative to standard methods in empirically relevant settings.They should thus be useful for practitioners. *(abstract via OpenAlex)*

Econometrica Econometrics & Economics · 发表于 2017

Research Design Meets Market Design: Using Centralized Assignment for Impact Evaluation

Atila Abdulkadiroğlu, Joshua D. Angrist, Yusuke Narita, Parag A. Pathak

摘要

A growing number of school districts use centralized assignment mechanisms to allocate school seats in a manner that reflects student preferences and school priorities.Many of these assignment schemes use lotteries to ration seats when schools are oversubscribed.The resulting random assignment opens the door to credible quasi-experimental research designs for the evaluation of school effectiveness.Yet the question of how best to separate the lottery-generated variation integral to such designs from non-random preferences and priorities remains open.This paper develops easily-implemented empirical strategies that fully exploit the random assignment embedded in a wide class of mechanisms, while also revealing why seats are randomized at one school but not another.We use these methods to evaluate charter schools in Denver, one of a growing number of districts that combine charter and traditional public schools in a unified assignment system.The resulting estimates show large achievement gains from charter school attendance.Our approach generates efficiency gains over ad hoc methods, such as those that focus on schools ranked first, while also identifying a more representative average causal effect.We also show how to use centralized assignment mechanisms to identify causal effects in models with multiple school sectors. *(abstract via OpenAlex)*

JMLR Machine Learning · 发表于 2017

Relational Reinforcement Learning for Planning with Exogenous Effects

David Martínez, Guillem Alenyà, Tony Ribeiro, Katsumi Inoue, Carme Torras

摘要

Probabilistic planners have improved recently to the point that they can solve difficult tasks with complex and expressive models. In contrast, learners cannot tackle yet the expressive models that planners do, which forces complex models to be mostly handcrafted. We propose a new learning approach that can learn relational probabilistic models with both action effects and exogenous effects. The proposed learning approach combines a multi-valued variant of inductive logic programming for the generation of candidate models, with an optimization method to select the best set of planning operators to model a problem. We also show how to combine this learner with reinforcement learning algorithms to solve complete problems. Finally, experimental validation is provided that shows improvements over previous work in both simulation and a robotic task. The robotic task involves a dynamic scenario with several agents where a manipulator robot has to clear the tableware on a table. We show that the exogenous effects learned by our approach allowed the robot to clear the table in a more efficient way.

JMLR High-Dimensional Spatial & Time Series Econometrics & Economics · 发表于 2017

Regularized Estimation and Testing for High-Dimensional Multi-Block Vector-Autoregressive Models

Jiahe Lin, George Michailidis

摘要

Dynamical systems comprising of multiple components that can be partitioned into distinct blocks originate in many scientific areas. A pertinent example is the interactions between financial assets and selected macroeconomic indicators, which has been studied at aggregate level---e.g. a stock index and an employment index---extensively in the macroeconomics literature. A key shortcoming of this approach is that it ignores potential influences from other related components (e.g. Gross Domestic Product) that may impact the system's dynamics and structure and thus produces incorrect results. To mitigate this issue, we consider a multi-block linear dynamical system with Granger-causal ordering between blocks, wherein the blocks' temporal dynamics are described by vector autoregressive processes and are influenced by blocks higher in the system hierarchy. We derive the maximum likelihood estimator for the posited model for Gaussian data in the high- dimensional setting based on appropriate regularization schemes for the parameters of the block components. To optimize the underlying non-convex likelihood function, we develop an iterative algorithm with convergence guarantees. We establish theoretical properties of the maximum likelihood estimates, leveraging the decomposability of the regularizers and a careful analysis of the iterates. Finally, we develop testing procedures for the null hypothesis of whether a block Granger-causes another block of variables. The performance of the model and the testing procedures are evaluated on synthetic data, and illustrated on a data set involving log-returns of the US S&P100 component stocks and key macroeconomic variables for the 2001--16 period.

JMLR Other · 发表于 2017

Regularization and the small-ball method II: complexity dependent error rates

Guillaume Lecué, Shahar Mendelson

摘要

We study estimation properties of regularized procedures of the form $\hat f \in\arg\min_{f\in F}\Big(\frac{1}{N}\sum_{i=1}^N\big(Y_i-f(X_i)\big)^2+\lambda \Psi(f)\Big)$ for a convex class of functions $F$, regularization function $\Psi(\cdot)$ and some well chosen regularization parameter $\lambda$, where the given data is an independent sample $(X_i, Y_i)_{i=1}^N$. We obtain bounds on the $L_2$ estimation error rate that depend on the complexity of the true model $F^*:=\{f\in F: \Psi(f)\leq\Psi(f^*)\}$, where $f^*\in\arg\min_{f\in F}\mathbb{E}(Y-f(X))^2$ and the $(X_i,Y_i)$'s are independent and distributed as $(X,Y)$. Our estimate holds under weak stochastic assumptions -- one of which being a small-ball condition satisfied by $F$ -- and for rather flexible choices of regularization functions $\Psi(\cdot)$. Moreover, the result holds in the learning theory framework: we do not assume any a-priori connection between the output $Y$ and the input $X$. As a proof of concept, we apply our general estimation bound to various choices of $\Psi$, for example, the $\ell_p$ and $S_p$-norms (for $p\geq1$), weak-$\ell_p$, atomic norms, max- norm and SLOPE. In many cases, the estimation rate almost coincides with the minimax rate in the class $F^*$.

JMLR Other · 发表于 2017

Refinery: An Open Source Topic Modeling Web Platform

Daeil Kim, Benjamin F. Swanson, Michael C. Hughes, Erik B. Sudderth

摘要

We introduce Refinery, an open source platform for exploring large text document collections with topic models. Refinery is a standalone web application driven by a graphical interface, so it is usable by those without machine learning or programming expertise. Users can interactively organize articles by topic and also refine this organization with phrase-level analysis. Under the hood, we train Bayesian nonparametric topic models that can adapt model complexity to the provided data with scalable learning algorithms. The project website contains Python code and further documentation.

Econometrica Econometrics & Economics · 发表于 2017

Recursive Equilibria in Dynamic Economies With Stochastic Production

Johannes Brumm, Dominika Kryczka, Felix Kubler

摘要

In this paper we prove the existence of recursive equilibria in stochastic production economies with infinitely lived agents and incomplete financial markets. We consider a general dynamic model with several commodities, which encompasses heterogeneous agent versions of both the Lucas asset pricing model and the stochastic neo-classical growth model as special cases. Our main assumption is that there are atomless shocks to fundamentals that have a purely transitory component and a component that does not depend on last period's shocks directly. *(abstract via OpenAlex)*

JMLR High-Dimensional · 发表于 2017

Recovering PCA and Sparse PCA via Hybrid-(l1,l2) Sparse Sampling of Data Elements

Abhisek Kundu, Petros Drineas, Malik Magdon-Ismail

摘要

This paper addresses how well we can recover a data matrix when only given a few of its elements. We present a randomized algorithm that element-wise sparsifies the data, retaining only a few of its entries. Our new algorithm independently samples the data using probabilities that depend on both squares ($\ell_2$ sampling) and absolute values ($\ell_1$ sampling) of the entries. We prove that this hybrid algorithm ($i$) achieves a near-PCA reconstruction of the data, and ($ii$) recovers sparse principal components of the data, from a sketch formed by a sublinear sample size. Hybrid-($\ell_1,\ell_2$) inherits the $\ell_2$-ability to sample the important elements, as well as the regularization properties of $\ell_1$ sampling, and maintains strictly better quality than either $\ell_1$ or $\ell_2$ on their own. Extensive experimental results on synthetic, image, text, biological, and financial data show that not only are we able to recover PCA and sparse PCA from incomplete data, but we can speed up such computations significantly using our sparse sketch .

JMLR Networks & Graphs · 发表于 2017

Reconstructing Undirected Graphs from Eigenspaces

Yohann De Castro, Thibault Espinasse, Paul Rochet

摘要

We aim at recovering the weighted adjacency matrix $\mathsf{W}$ of an undirected graph from a perturbed version of its eigenspaces. This situation arises for instance when working with stationary signals on graphs or Markov chains observed at random times. Our approach relies on minimizing a cost function based on the Frobenius norm of the commutator $\mathsf{A} \mathsf{B}-\mathsf{B} \mathsf{A}$ between symmetric matrices $\mathsf{A}$ and $\mathsf{B}$. We describe a particular framework in which we have access to an estimation of the eigenspaces and provide support selection procedures from theoretical and practical points of view. In the Erdős-Rényi model on $N$ vertices with no self-loops, we show that identifiability (i.e., the ability to reconstruct $\mathsf{W}$ from the knowledge of its eigenspaces) follows a sharp phase transition on the expected number of edges with threshold function $N\log N/2$. Simulated and real life numerical experiments assert our methodology.

Econometrica Other · 发表于 2017

Rational Inattention Dynamics: Inertia and Delay in Decision-Making

Jakub Steiner, Colin Stewart, Filip Matějka

摘要

We solve a general class of dynamic rational-inattention problems in which an agent repeatedly acquires costly information about an evolving state and selects actions. The solution resembles the choice rule in a dynamic logit model, but it is biased towards an optimal default rule that does not depend on the realized state. We apply the general solution to the study of (i) the sunk-cost fallacy; (ii) inertia in actions leading to lagged adjustments to shocks; and (iii) the tradeoff between accuracy and delay in decision-making. *(abstract via OpenAlex)*

JMLR Other · 发表于 2017

Rank Determination for Low-Rank Data Completion

Morteza Ashraphijuo, Xiaodong Wang, Vaneet Aggarwal

摘要

Recently, fundamental conditions on the sampling patterns have been obtained for finite completability of low-rank matrices or tensors given the corresponding ranks. In this paper, we consider the scenario where the rank is not given and we aim to approximate the unknown rank based on the location of sampled entries and some given completion. We consider a number of data models, including single-view matrix, multi-view matrix, CP tensor, tensor-train tensor and Tucker tensor. For each of these data models, we provide an upper bound on the rank when an arbitrary low-rank completion is given. We characterize these bounds both deterministically, i.e., with probability one given that the sampling pattern satisfies certain combinatorial properties, and probabilistically, i.e., with high probability given that the sampling probability is above some threshold. Moreover, for both single-view matrix and CP tensor, we are able to show that the obtained upper bound is exactly equal to the unknown rank if the lowest-rank completion is given. Furthermore, we provide numerical experiments for the case of single-view matrix, where we use nuclear norm minimization to find a low-rank completion of the sampled data and we observe that in most of the cases the proposed upper bound on the rank is equal to the true rank.

Econometrica Testing & Uncertainty · 发表于 2017

Randomization Tests Under an Approximate Symmetry Assumption

Ivan A. Canay, Joseph P. Romano, Azeem M. Shaikh

摘要

This paper develops a theory of randomization tests under an approximate symmetry as-sumption. Randomization tests provide a general means of constructing tests that control size in finite samples whenever the distribution of the observed data exhibits symmetry under the null hypothesis. Here, by exhibits symmetry we mean that the distribution remains invariant under a group of transformations. In this paper, we provide conditions under which the same construction can be used to construct tests that asymptotically control the probability of a false rejection whenever the distribution of the observed data exhibits approximate symmetry in the sense that the limiting distribution of a function of the data exhibits symmetry under the null hypothesis. An important application of this idea is in settings where the data may be grouped into a fixed number of “clusters ” with a large number of observations within each cluster. In such settings, we show that the distribution of the observed data satisfies our ap-proximate symmetry requirement under weak assumptions. In particular, our results allow for the clusters to be heterogeneous and also have dependence not only within each cluster, but also across clusters. This approach enjoys several advantages over other approaches in these settings. Among other things, it leads to a test that is asymptotically similar, which, as shown in a simulation study, translates into improved power at many alternatives. Finally, we use our results to revisit the analysis of Angrist and Lavy (2009), who examine the impact of a cash award on exam performance for low-achievement students in Israel. *(abstract via OpenAlex)*

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 页