Practical or scientific considerations often lead to selecting a subset of parameters as ``important.'' Inferences about those parameters often are based on the same data used to select them in the first place. That can make the reported uncertainties deceptively optimistic: confidence intervals that ignore selection generally have less than their nominal coverage probability. Controlling the probability that one or more intervals for selected parameters do not cover---the ``simultaneous over the selected'' (SoS) error rate---is crucial in many scientific problems. Intervals that control the SoS error rate can be constructed in ways that take advantage of knowledge of the selection rule. We construct SoS-controlling confidence intervals for parameters deemed the most ``important'' $k$ of $m$ shift parameters because they are estimated (by independent estimators) to be the largest. The new intervals improve substantially over \v{S}id\'{a}k intervals when $k$ is small compared to $m$, and approach the standard Bonferroni-corrected intervals when $k \approx m$. Standard, unadjusted confidence intervals for location parameters have the correct coverage probability for $k=1$, $m=2$ if, when the true parameters are zero, the estimators are exchangeable and symmetric.

Real world observational data, together with causal inference, allow the estimation of causal effects when randomized controlled trials are not available. To be accepted into practice, such predictive models must be validated for the dataset at hand, and thus require a comprehensive evaluation toolkit, as introduced here. Since effect estimation cannot be evaluated directly, we turn to evaluating the various observable properties of causal inference, namely the observed outcome and treatment assignment. We developed a toolkit that expands established machine learning evaluation methods and adds several causal-specific ones. Evaluations can be applied in cross-validation, in a train-test scheme, or on the training data. Multiple causal inference methods are implemented within the toolkit in a way that allows modular use of the underlying machine learning models. Thus, the toolkit is agnostic to the machine learning model that is used. We showcase our approach using a rheumatoid arthritis cohort (consisting of about 120K patients) extracted from the IBM MarketScan(R) Research Database. We introduce an iterative pipeline of data definition, model definition, and model evaluation. Using this pipeline, we demonstrate how each of the evaluation components helps drive model selection and refinement of data extraction criteria in a way that provides more reproducible results and ensures that the causal question is answerable with available data. Furthermore, we show how the evaluation toolkit can be used to ensure that performance is maintained when applied to subsets of the data, thus allowing exploration of questions that move towards personalized medicine.

We take a Hamiltonian-based perspective to generalize Nesterov's accelerated gradient descent and Polyak's heavy ball method to a broad class of momentum methods in the setting of (possibly) constrained minimization in Banach spaces. Our perspective leads to a generic and unifying non-asymptotic analysis of convergence of these methods in both the function value (in the setting of convex optimization) and in the norm of the gradient (in the setting of unconstrained, possibly nonconvex, optimization). The convergence analysis is intuitive and based on the conserved quantities of the time-dependent Hamiltonian that we introduce and that produces generalized momentum methods as its equations of motion.

We propose a differentiable nonparametric algorithm, the Delaunay triangulation learner (DTL), to solve the functional approximation problem on the basis of a $p$-dimensional feature space. By conducting the Delaunay triangulation algorithm on the data points, the DTL partitions the feature space into a series of $p$-dimensional simplices in a geometrically optimal way, and fits a linear model within each simplex. We study its theoretical properties by exploring the geometric properties of the Delaunay triangulation, and compare its performance with other statistical learners in numerical studies.

In this paper, we consider derivative free optimization problems, where the objective function is smooth but is computed with some amount of noise, the function evaluations are expensive and no derivative information is available. We are motivated by policy optimization problems in reinforcement learning that have recently become popular [Choromaski et al. 2018; Fazel et al. 2018; Salimans et al. 2016], and that can be formulated as derivative free optimization problems with the aforementioned characteristics. In each of these works some approximation of the gradient is constructed and a (stochastic) gradient method is applied. In [Salimans et al. 2016] the gradient information is aggregated along Gaussian directions, while in [Choromaski et al. 2018] it is computed along orthogonal direction. We provide a convergence rate analysis for a first-order line search method, similar to the ones used in the literature, and derive the conditions on the gradient approximations that ensure this convergence. We then demonstrate via rigorous analysis of the variance and by numerical comparisons on reinforcement learning tasks that the Gaussian sampling method used in [Salimans et al. 2016] is significantly inferior to the orthogonal sampling used in [Choromaski et al. 2018] as well as more general interpolation methods.

Uncertainty quantification has been a core of the statistical machine learning, but its computational bottleneck has been a serious challenge for both Bayesians and frequentists. We propose a model-based framework in quantifying uncertainty, called predictive-matching Generative Parameter Sampler (GPS). This procedure considers an Uncertainty Quantification (UQ) distribution, on the targeted parameter, which matches the corresponding predictive distribution to the observed data. This framework adopts a hierarchical modeling perspective such that each observation is modeled by an individual parameter. This individual parameterization permits the resulting inference to be computationally scalable and robust to outliers. Our approach is illustrated for linear models, Poisson processes, and deep neural networks for classification. The results show that the GPS is successful in providing uncertainty quantification as well as additional flexibility beyond what is allowed by classical statistical procedures under the postulated statistical models.

This paper examines the assumptions of the derived equivalence between dropout noise injection and $L_2$ regularisation for logistic regression with negative log loss. We show that the approximation method is based on a divergent Taylor expansion, making, subsequent work using this approximation to compare the dropout trained logistic regression model with standard regularisers unfortunately ill-founded to date. Moreover, the approximation approach is shown to be invalid using any robust constraints. We show how this finding extends to general neural network topologies that use a cross-entropy prediction layer.

Recent results in Reinforcement Learning (RL) have shown that agents with limited training environments are susceptible to a large amount of overfitting across many domains. A key challenge for RL generalization is to quantitatively explain the effects of changing parameters on testing performance. Such parameters include architecture, regularization, and RL-dependent variables such as discount factor and action stochasticity. We provide empirical results that show complex and interdependent relationships between hyperparameters and generalization. We further show that several empirical metrics such as gradient cosine similarity and trajectory-dependent metrics serve to provide intuition towards these results.

We construct and analyze active learning algorithms for the problem of binary classification with abstention. We consider three abstention settings: \emph{fixed-cost} and two variants of \emph{bounded-rate} abstention, and for each of them propose an active learning algorithm. All the proposed algorithms can work in the most commonly used active learning models, i.e., \emph{membership-query}, \emph{pool-based}, and \emph{stream-based} sampling. We obtain upper-bounds on the excess risk of our algorithms in a general non-parametric framework and establish their minimax near-optimality by deriving matching lower-bounds. Since our algorithms rely on the knowledge of some smoothness parameters of the regression function, we then describe a new strategy to adapt to these unknown parameters in a data-driven manner. Since the worst case computational complexity of our proposed algorithms increases exponentially with the dimension of the input space, we conclude the paper with a computationally efficient variant of our algorithm whose computational complexity has a polynomial dependence over a smaller but rich class of learning problems.

Recent applications that arise in machine learning have surged significant interest in solving min-max saddle point games. This problem has been extensively studied in the convex-concave regime for which a global equilibrium solution can be computed efficiently. In this paper, we study the problem in the non-convex regime and show that an \varepsilon--first order stationary point of the game can be computed when one of the player's objective can be optimized to global optimality efficiently. In particular, we first consider the case where the objective of one of the players satisfies the Polyak-{\L}ojasiewicz (PL) condition. For such a game, we show that a simple multi-step gradient descent-ascent algorithm finds an \varepsilon--first order stationary point of the problem in \widetilde{\mathcal{O}}(\varepsilon^{-2}) iterations. Then we show that our framework can also be applied to the case where the objective of the ``max-player'' is concave. In this case, we propose a multi-step gradient descent-ascent algorithm that finds an \varepsilon--first order stationary point of the game in \widetilde{\cal O}(\varepsilon^{-3.5}) iterations, which is the best known rate in the literature. We applied our algorithm to a fair classification problem of Fashion-MNIST dataset and observed that the proposed algorithm results in smoother training and better generalization.