Markov models are often used to capture the temporal patterns of sequential data for statistical learning applications. While the Hidden Markov modeling-based learning mechanisms are well studied in literature, we analyze a symbolic-dynamics inspired approach. Under this umbrella, Markov modeling of time-series data consists of two major steps -- discretization of continuous attributes followed by estimating the size of temporal memory of the discretized sequence. These two steps are critical for the accurate and concise representation of time-series data in the discrete space. Discretization governs the information content of the resultant discretized sequence. On the other hand, memory estimation of the symbolic sequence helps to extract the predictive patterns in the discretized data. Clearly, the effectiveness of signal representation as a discrete Markov process depends on both these steps. In this paper, we will review the different techniques for discretization and memory estimation for discrete stochastic processes. In particular, we will focus on the individual problems of discretization and order estimation for discrete stochastic process. We will present some results from literature on partitioning from dynamical systems theory and order estimation using concepts of information theory and statistical learning. The paper also presents some related problem formulations which will be useful for machine learning and statistical learning application using the symbolic framework of data analysis. We present some results of statistical analysis of a complex thermoacoustic instability phenomenon during lean-premixed combustion in jet-turbine engines using the proposed Markov modeling method.

Open Source Software projects add labels to open issues to help contributors choose tasks. However, manually labeling issues is time-consuming and error-prone. Current automatic approaches for creating labels are mostly limited to classifying issues as a bug/non-bug. In this paper, we investigate the feasibility and relevance of labeling issues with the domain of the APIs required to complete the tasks. We leverage the issues' description and the project history to build prediction models, which resulted in precision up to 82% and recall up to 97.8%. We also ran a user study (n=74) to assess these labels' relevancy to potential contributors. The results show that the labels were useful to participants in choosing tasks, and the API-domain labels were selected more often than the existing architecture-based labels. Our results can inspire the creation of tools to automatically label issues, helping developers to find tasks that better match their skills.

We propose a novel information bottleneck (IB) method named Drop-Bottleneck, which discretely drops features that are irrelevant to the target variable. Drop-Bottleneck not only enjoys a simple and tractable compression objective but also additionally provides a deterministic compressed representation of the input variable, which is useful for inference tasks that require consistent representation. Moreover, it can jointly learn a feature extractor and select features considering each feature dimension's relevance to the target task, which is unattainable by most neural network-based IB methods. We propose an exploration method based on Drop-Bottleneck for reinforcement learning tasks. In a multitude of noisy and reward sparse maze navigation tasks in VizDoom (Kempka et al., 2016) and DM-Lab (Beattie et al., 2016), our exploration method achieves state-of-the-art performance. As a new IB framework, we demonstrate that Drop-Bottleneck outperforms Variational Information Bottleneck (VIB) (Alemi et al., 2017) in multiple aspects including adversarial robustness and dimensionality reduction. Data with noise or task-irrelevant information easily harm the training of a model; for instance, the noisy-TV problem (Burda et al., 2019a) is one of well-known such phenomena in reinforcement learning. If observations from the environment are modified to contain a TV screen, which changes its channel randomly based on the agent's actions, the performance of curiosity-based exploration methods dramatically degrades (Burda et al., 2019a;b; Kim et al., 2019; Savinov et al., 2019). The information bottleneck (IB) theory (Tishby et al., 2000; Tishby & Zaslavsky, 2015) provides a framework for dealing with such task-irrelevant information, and has been actively adopted to exploration in reinforcement learning (Kim et al., 2019; Igl et al., 2019). For an input variable X and a target variable Y, the IB theory introduces another variable Z, which is a compressed representation of X.

Modern applications of survival analysis increasingly involve time-dependent covariates, which constitute a form of functional data. Learning from functional data generally involves repeated evaluations of time integrals which is numerically expensive. In this work we propose a lightweight data preprocessing step that transforms functional data into nonfunctional data. Boosting implementations for nonfunctional data can then be used, whereby the required numerical integration comes for free as part of the training phase. We use this to develop BoXHED 2.0, a quantum leap over the tree-boosted hazard package BoXHED 1.0. BoXHED 2.0 extends BoXHED 1.0 to Aalen's multiplicative intensity model, which covers censoring schemes far beyond right-censoring and also supports recurrent events data. It is also massively scalable because of preprocessing and also because it borrows from the core components of XGBoost. BoXHED 2.0 supports the use of GPUs and multicore CPUs, and is available from GitHub: www.github.com/BoXHED.

We introduce a simple, rigorous, and unified framework for solving nonlinear partial differential equations (PDEs), and for solving inverse problems (IPs) involving the identification of parameters in PDEs, using the framework of Gaussian processes. The proposed approach (1) provides a natural generalization of collocation kernel methods to nonlinear PDEs and IPs, (2) has guaranteed convergence with a path to compute error bounds in the PDE setting, and (3) inherits the state-of-the-art computational complexity of linear solvers for dense kernel matrices. The main idea of our method is to approximate the solution of a given PDE with a MAP estimator of a Gaussian process given the observation of the PDE at a finite number of collocation points. Although this optimization problem is infinite-dimensional, it can be reduced to a finite-dimensional one by introducing additional variables corresponding to the values of the derivatives of the solution at collocation points; this generalizes the representer theorem arising in Gaussian process regression. The reduced optimization problem has a quadratic loss and nonlinear constraints, and it is in turn solved with a variant of the Gauss-Newton method. The resulting algorithm (a) can be interpreted as solving successive linearizations of the nonlinear PDE, and (b) is found in practice to converge in a small number (two to ten) of iterations in experiments conducted on a range of PDEs. For IPs, while the traditional approach has been to iterate between the identifications of parameters in the PDE and the numerical approximation of its solution, our algorithm tackles both simultaneously. Experiments on nonlinear elliptic PDEs, Burgers' equation, a regularized Eikonal equation, and an IP for permeability identification in Darcy flow illustrate the efficacy and scope of our framework.

Logistic regression remains one of the most widely used tools in applied statistics, machine learning and data science. Practical datasets often have a substantial number of features $d$ relative to the sample size $n$. In these cases, the logistic regression maximum likelihood estimator (MLE) is biased, and its standard large-sample approximation is poor. In this paper, we develop an improved method for debiasing predictions and estimating frequentist uncertainty for such datasets. We build on recent work characterizing the asymptotic statistical behavior of the MLE in the regime where the aspect ratio $d / n$, instead of the number of features $d$, remains fixed as $n$ grows. In principle, this approximation facilitates bias and uncertainty corrections, but in practice, these corrections require an estimate of the signal strength of the predictors. Our main contribution is SLOE, an estimator of the signal strength with convergence guarantees that reduces the computation time of estimation and inference by orders of magnitude. The bias correction that this facilitates also reduces the variance of the predictions, yielding narrower confidence intervals with higher (valid) coverage of the true underlying probabilities and parameters. We provide an open source package for this method, available at https://github.com/google-research/sloe-logistic.

Data depth is a non parametric statistical tool that measures centrality of any element $x\in\mathbb{R}^d$ with respect to (w.r.t.) a probability distribution or a data set. It is a natural median-oriented extension of the cumulative distribution function (cdf) to the multivariate case. Consequently, its upper level sets -- the depth-trimmed regions -- give rise to a definition of multivariate quantiles. In this work, we propose two new pseudo-metrics between continuous probability measures based on data depth and its associated central regions. The first one is constructed as the Lp-distance between data depth w.r.t. each distribution while the second one relies on the Hausdorff distance between their quantile regions. It can further be seen as an original way to extend the one-dimensional formulae of the Wasserstein distance, which involves quantiles and cdfs, to the multivariate space. After discussing the properties of these pseudo-metrics and providing conditions under which they define a distance, we highlight similarities with the Wasserstein distance. Interestingly, the derived non-asymptotic bounds show that in contrast to the Wasserstein distance, the proposed pseudo-metrics do not suffer from the curse of dimensionality. Moreover, based on the support function of a convex body, we propose an efficient approximation possessing linear time complexity w.r.t. the size of the data set and its dimension. The quality of this approximation as well as the performance of the proposed approach are illustrated in experiments. Furthermore, by construction the regions-based pseudo-metric appears to be robust w.r.t. both outliers and heavy tails, a behavior witnessed in the numerical experiments.

There is an increasing realization that algorithmic inductive biases are central in preventing overfitting; empirically, we often see a benign overfitting phenomenon in overparameterized settings for natural learning algorithms, such as stochastic gradient descent (SGD), where little to no explicit regularization has been employed. This work considers this issue in arguably the most basic setting: constant-stepsize SGD (with iterate averaging) for linear regression in the overparameterized regime. Our main result provides a sharp excess risk bound, stated in terms of the full eigenspectrum of the data covariance matrix, that reveals a bias-variance decomposition characterizing when generalization is possible: (i) the variance bound is characterized in terms of an effective dimension (specific for SGD) and (ii) the bias bound provides a sharp geometric characterization in terms of the location of the initial iterate (and how it aligns with the data covariance matrix). We reflect on a number of notable differences between the algorithmic regularization afforded by (unregularized) SGD in comparison to ordinary least squares (minimum-norm interpolation) and ridge regression.

Equal-weighted portfolios are one of the most important strategies in portfolio management. They are portfolios with weights equally distributed across the selected securities in the long and/or short positions. In academic research, numerous studies have suggested that equal-weighted portfolios have a better out-of-sample performance than other portfolios (e.g., Jobson and Korkie 1981; James 2003; DeMiguel et al. 2007). Michaud (1989) and DeMiguel et al. (2007) argued that, the equal-weighted strategies do not suffer from the estimation error of the covariance matrix, which is vulnerable to outliers (Tu and Zhou, 2011). In industry, equal-weighted portfolios are popular across portfolio management in practice, particularly in the hedge funds. The MSCI has issued many equal-weighted indexes, which are "some of the oldest and best-known factor strategies that have aimed to identify specific characteristics of stocks generating excess return"

Despite the strong theoretical guarantees that variance-reduced finite-sum optimization algorithms enjoy, their applicability remains limited to cases where the memory overhead they introduce (SAG/SAGA), or the periodic full gradient computation they require (SVRG/SARAH) are manageable. A promising approach to achieving variance reduction while avoiding these drawbacks is the use of importance sampling instead of control variates. While many such methods have been proposed in the literature, directly proving that they improve the convergence of the resulting optimization algorithm has remained elusive. In this work, we propose an importance-sampling-based algorithm we call SRG (stochastic reweighted gradient). We analyze the convergence of SRG in the strongly-convex case and show that, while it does not recover the linear rate of control variates methods, it provably outperforms SGD. We pay particular attention to the time and memory overhead of our proposed method, and design a specialized red-black tree allowing its efficient implementation. Finally, we present empirical results to support our findings.