<?xml version="1.0" encoding="UTF-8"?>
 <rdf:RDF xmlns="http://purl.org/rss/1.0/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://web.resource.org/cc/" xmlns:syn="http://purl.org/rss/1.0/modules/syndication/" xmlns:admin="http://webns.net/mvcb/">
  <channel rdf:about="http://pinboard.in">
    <title>Pinboard (cshalizi)</title>
    <link>https://pinboard.in/u:cshalizi/public/</link>
    <description>recent bookmarks from cshalizi</description>
    <items>
      <rdf:Seq>	<rdf:li rdf:resource="https://projecteuclid.org/journals/annals-of-statistics/volume-49/issue-3/Learning-models-with-uniform-performance-via-distributionally-robust-optimization/10.1214/20-AOS2004.short"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2103.04668"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2105.03067"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2101.06309"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/1905.04281"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2012.02914"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2011.14999"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2011.10529"/>
	<rdf:li rdf:resource="https://www.annualreviews.org/doi/abs/10.1146/annurev-control-091219-012549"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2003.00688"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/1908.05659"/>
	<rdf:li rdf:resource="http://arxiv.org/abs/1402.6118"/>
	<rdf:li rdf:resource="http://arxiv.org/abs/1206.4628"/>
	<rdf:li rdf:resource="http://jmlr.csail.mit.edu/papers/v9/debruyne08a.html"/>
	<rdf:li rdf:resource="http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism_25.html"/>
	<rdf:li rdf:resource="http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism-iv.html"/>
	<rdf:li rdf:resource="http://projecteuclid.org/DPubS?service=UI&amp;version=1.0&amp;verb=Display&amp;handle=euclid.lnms/1249305323"/>
	<rdf:li rdf:resource="http://arxiv.org/abs/0804.4714"/>
	<rdf:li rdf:resource="http://bookstaber.com/rick/OnTheOptimalityOfCoarseBehavior.pdf"/>
	<rdf:li rdf:resource="http://arxiv.org/abs/0710.4269"/>
      </rdf:Seq>
    </items>
  </channel><item rdf:about="https://projecteuclid.org/journals/annals-of-statistics/volume-49/issue-3/Learning-models-with-uniform-performance-via-distributionally-robust-optimization/10.1214/20-AOS2004.short">
    <title>Learning models with uniform performance via distributionally robust optimization</title>
    <dc:date>2021-08-09T16:24:20+00:00</dc:date>
    <link>https://projecteuclid.org/journals/annals-of-statistics/volume-49/issue-3/Learning-models-with-uniform-performance-via-distributionally-robust-optimization/10.1214/20-AOS2004.short</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["A common goal in statistics and machine learning is to learn models that can perform well against distributional shifts, such as latent heterogeneous subpopulations, unknown covariate shifts or unmodeled temporal effects. We develop and analyze a distributionally robust stochastic optimization (DRO) framework that learns a model providing good performance against perturbations to the data-generating distribution. We give a convex formulation for the problem, providing several convergence guarantees. We prove finite-sample minimax upper and lower bounds, showing that distributional robustness sometimes comes at a cost in convergence rates. We give limit theorems for the learned parameters, where we fully specify the limiting distribution so that confidence intervals can be computed. On real tasks including generalizing to unknown subpopulations, fine-grained recognition and providing good tail performance, the distributionally robust approach often exhibits improved performance."]]></description>
<dc:subject>to:NB optimization statistics prediction learning_theory non-stationarity robustness re:codename:one_law_for_the_lion_and_ox_is_oppression</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:064040e8a17c/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:prediction"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:non-stationarity"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:codename:one_law_for_the_lion_and_ox_is_oppression"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2103.04668">
    <title>[2103.04668] The distance backbone of complex networks</title>
    <dc:date>2021-05-14T01:54:54+00:00</dc:date>
    <link>https://arxiv.org/abs/2103.04668</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Redundancy needs more precise characterization as it is a major factor in the evolution and robustness of networks of multivariate interactions. We investigate the complexity of such interactions by inferring a connection transitivity that includes all possible measures of path length for weighted graphs. The result, without breaking the graph into smaller components, is a distance backbone subgraph sufficient to compute all shortest paths. This is important for understanding the dynamics of spread and communication phenomena in real-world networks. The general methodology we formally derive yields a principled graph reduction technique and provides a finer characterization of the triangular geometry of all edges -- those that contribute to shortest paths and those that do not but are involved in other network phenomena. We demonstrate that the distance backbone is very small in large networks across domains ranging from air traffic to the human brain connectome, revealing that network robustness to attacks and failures seems to stem from surprisingly vast amounts of redundancy."]]></description>
<dc:subject>to:NB network_data_analysis robustness graph_theory rocha.luis_m.</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:112f46a8d854/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:network_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:graph_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:rocha.luis_m."/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2105.03067">
    <title>[2105.03067] The $r$-value: evaluating stability with respect to distributional shifts</title>
    <dc:date>2021-05-10T22:50:22+00:00</dc:date>
    <link>https://arxiv.org/abs/2105.03067</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Common statistical measures of uncertainty like p-values and confidence intervals quantify the uncertainty due to sampling, that is, the uncertainty due to not observing the full population. In practice, populations change between locations and across time. This makes it difficult to gather knowledge that transfers across data sets. We propose a measure of uncertainty that quantifies the distributional uncertainty of a statistical estimand with respect to Kullback-Liebler divergence, that is, the sensitivity of the parameter under general distributional perturbations within a Kullback-Liebler divergence ball. If the signal-to-noise ratio is small, distributional uncertainty is a monotonous transformation of the signal-to-noise ratio. In general, however, it is a different concept and corresponds to a different research question. Further, we propose measures to estimate the stability of parameters with respect to directional or variable-specific shifts. We also demonstrate how the measure of distributional uncertainty can be used to prioritize data collection for better estimation of statistical parameters under shifted distribution. We evaluate the performance of the proposed measure in simulations and real data and show that it can elucidate the distributional (in-)stability of an estimator with respect to certain shifts and give more accurate estimates of parameters under shifted distribution only requiring to collect limited information from the shifted distribution."]]></description>
<dc:subject>to:NB information_theory robustness statistics misspecification sensitivity_analysis</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:c7ec017eda3a/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:information_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:sensitivity_analysis"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2101.06309">
    <title>[2101.06309] Fundamental Tradeoffs in Distributionally Adversarial Training</title>
    <dc:date>2021-01-19T18:33:29+00:00</dc:date>
    <link>https://arxiv.org/abs/2101.06309</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Adversarial training is among the most effective techniques to improve the robustness of models against adversarial perturbations. However, the full effect of this approach on models is not well understood. For example, while adversarial training can reduce the adversarial risk (prediction error against an adversary), it sometimes increase standard risk (generalization error when there is no adversary). Even more, such behavior is impacted by various elements of the learning problem, including the size and quality of training data, specific forms of adversarial perturbations in the input, model overparameterization, and adversary's power, among others. In this paper, we focus on \emph{distribution perturbing} adversary framework wherein the adversary can change the test distribution within a neighborhood of the training data distribution. The neighborhood is defined via Wasserstein distance between distributions and the radius of the neighborhood is a measure of adversary's manipulative power. We study the tradeoff between standard risk and adversarial risk and derive the Pareto-optimal tradeoff, achievable over specific classes of models, in the infinite data limit with features dimension kept fixed. We consider three learning settings: 1) Regression with the class of linear models; 2) Binary classification under the Gaussian mixtures data model, with the class of linear classifiers; 3) Regression with the class of random features model (which can be equivalently represented as two-layer neural network with random first-layer weights). We show that a tradeoff between standard and adversarial risk is manifested in all three settings. We further characterize the Pareto-optimal tradeoff curves and discuss how a variety of factors, such as features correlation, adversary's power or the width of two-layer neural network would affect this tradeoff."

--- Ain't this just robustness to mis-specification?]]></description>
<dc:subject>to:NB statistics misspecification regression classifiers adversarial_examples robustness color_me_skeptical</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:b9c81909cdee/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:classifiers"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:adversarial_examples"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:color_me_skeptical"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/1905.04281">
    <title>[1905.04281] Robust high dimensional learning for Lipschitz and convex losses</title>
    <dc:date>2021-01-07T21:44:49+00:00</dc:date>
    <link>https://arxiv.org/abs/1905.04281</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We establish risk bounds for Regularized Empirical Risk Minimizers (RERM) when the loss is Lipschitz and convex and the regularization function is a norm. In a first part, we obtain these results in the i.i.d. setup under subgaussian assumptions on the design. In a second part, a more general framework where the design might have heavier tails and data may be corrupted by outliers both in the design and the response variables is considered. In this situation, RERM performs poorly in general. We analyse an alternative procedure based on median-of-means principles and called minmax MOM. We show optimal subgaussian deviation rates for these estimators in the relaxed setting. The main results are meta-theorems allowing a wide-range of applications to various problems in learning theory. To show a non-exhaustive sample of these potential applications, it is applied to classification problems with logistic loss functions regularized by LASSO and SLOPE, to regression problems with Huber loss regularized by Group LASSO and Total Variation. Another advantage of the minmax MOM formulation is that it suggests a systematic way to slightly modify descent based algorithms used in high-dimensional statistics to make them robust to outliers. We illustrate this principle in a Simulations section where a minmax MOM version of classical proximal descent algorithms are turned into robust to outliers algorithms."]]></description>
<dc:subject>to:NB learning_theory high-dimensional_statistics robustness</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:fb2d3a9d111f/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:high-dimensional_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2012.02914">
    <title>[2012.02914] Robustness on Networks</title>
    <dc:date>2020-12-10T05:44:21+00:00</dc:date>
    <link>https://arxiv.org/abs/2012.02914</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We adopt the statistical framework on robustness proposed by Watson and Holmes in 2016 and then tackle the practical challenges that hinder its applicability to network models. The goal is to evaluate how the quality of an inference for a network feature degrades when the assumed model is misspecified. Decision theory methods aimed to identify model missespecification are applied in the context of network data with the goal of investigating the stability of optimal actions to perturbations to the assumed model. Here the modified versions of the model are contained within a well defined neighborhood of model space. Our main challenge is to combine stochastic optimization and graph limits tools to explore the model space. As a result, a method for robustness on exchangeable random networks is developed. Our approach is inspired by recent developments in the context of robustness and recent works in the robust control, macroeconomics and financial mathematics literature and more specifically and is based on the concept of graphon approximation through its empirical graphon."]]></description>
<dc:subject>to:NB network_data_analysis graphons robustness wolfe.patrick</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:df8016a5aee8/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:network_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:graphons"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:wolfe.patrick"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2011.14999">
    <title>[2011.14999] An Automatic Finite-Sample Robustness Metric: Can Dropping a Little Data Change Conclusions?</title>
    <dc:date>2020-12-03T16:13:23+00:00</dc:date>
    <link>https://arxiv.org/abs/2011.14999</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We propose a method to assess the sensitivity of econometric analyses to the removal of a small fraction of the sample. Analyzing all possible data subsets of a certain size is computationally prohibitive, so we provide a finite-sample metric to approximately compute the number (or fraction) of observations that has the greatest influence on a given result when dropped. We call our resulting metric the Approximate Maximum Influence Perturbation. Our approximation is automatically computable and works for common estimators (including OLS, IV, GMM, MLE, and variational Bayes). We provide explicit finite-sample error bounds on our approximation for linear and instrumental variables regressions. At minimal computational cost, our metric provides an exact finite-sample lower bound on sensitivity for any estimator, so any non-robustness our metric finds is conclusive. We demonstrate that the Approximate Maximum Influence Perturbation is driven by a low signal-to-noise ratio in the inference problem, is not reflected in standard errors, does not disappear asymptotically, and is not a product of misspecification. Several empirical applications show that even 2-parameter linear regression analyses of randomized trials can be highly sensitive. While we find some applications are robust, in others the sign of a treatment effect can be changed by dropping less than 1% of the sample even when standard errors are small."]]></description>
<dc:subject>to:NB statistics robustness estimation to_read linear_regression to_teach:linear_models to_teach:undergrad-ADA</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:bc72bd101639/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:linear_regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:linear_models"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:undergrad-ADA"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2011.10529">
    <title>[2011.10529] Computation capacities of a broad class of signaling networks are higher than their communication capacities</title>
    <dc:date>2020-11-23T17:36:13+00:00</dc:date>
    <link>https://arxiv.org/abs/2011.10529</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Due to structural and functional abnormalities or genetic variations and mutations, there may be dysfunctional molecules within an intracellular signaling network that do not allow the network to correctly regulate its output molecules, such as transcription factors. This disruption in signaling interrupts normal cellular functions and may eventually develop some pathological conditions. In this paper, computation capacity of signaling networks is introduced as a fundamental limit on signaling capability and performance of such networks. The computation capacity measures the maximum number of computable inputs, that is, the maximum number of input values for which the correct functional output values can be recovered from the erroneous network outputs, when the network contains some dysfunctional molecules. This contrasts with the conventional communication capacity that measures instead the maximum number of input values that can be correctly distinguished based on the erroneous network outputs.
"The computation capacity is higher than the communication capacity, if the network response function is not a one-to-one function of the input signals. By explicitly incorporating the effect of signaling errors that result in the network dysfunction, the computation capacity provides more information about the network and its malfunction. Two examples of signaling networks are studied here, one regulating caspase3 and another regulating NFkB, for which computation and communication capacities are analyzed. Higher computation capacities are observed for both networks. One biological implication of this finding is that signaling networks may have more capacity than that specified by the conventional communication capacity metric. The effect of feedback is also studied. In summary, this paper reports findings on a new fundamental feature of the signaling capability of cell signaling networks."]]></description>
<dc:subject>to:NB biochemical_networks robustness information_theory</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:b29e3e6e4e29/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:biochemical_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:information_theory"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.annualreviews.org/doi/abs/10.1146/annurev-control-091219-012549">
    <title>Network Effects on the Robustness of Dynamic Systems | Annual Review of Control, Robotics, and Autonomous Systems</title>
    <dc:date>2020-11-18T22:45:23+00:00</dc:date>
    <link>https://www.annualreviews.org/doi/abs/10.1146/annurev-control-091219-012549</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We review selected results related to the robustness of networked systems in finite and asymptotically large size regimes in static and dynamical settings. In the static setting, within the framework of flow over finite networks, we discuss the effect of physical constraints on robustness to loss in link capacities. In the dynamical setting, we review several settings in which small-gain-type analysis provides tight robustness guarantees for linear dynamics over finite networks toward worst-case and stochastic disturbances. We discuss network flow dynamic settings where nonlinear techniques facilitate understanding the effect, on robustness, of constraints on capacity and information, substituting information with control action, and cascading failure. We also contrast cascading failure with a representative contagion model. For asymptotically large networks, we discuss the role of network properties in connecting microscopic shocks to emergent macroscopic fluctuations under linear dynamics as well as for economic networks at equilibrium. Through this review, we aim to achieve two objectives: to highlight selected settings in which the role of the interconnectivity structure of a network in its robustness is well understood, and to highlight a few additional settings in which existing system-theoretic tools give tight robustness guarantees and that are also appropriate avenues for future network-theoretic investigations."]]></description>
<dc:subject>to:NB networks dynamical_systems control_theory_and_control_engineering robustness</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:4debb7827817/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:dynamical_systems"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:control_theory_and_control_engineering"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2003.00688">
    <title>[2003.00688] Out-of-Distribution Generalization via Risk Extrapolation (REx)</title>
    <dc:date>2020-04-14T17:30:51+00:00</dc:date>
    <link>https://arxiv.org/abs/2003.00688</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Generalizing outside of the training distribution is an open challenge for current machine learning systems. A weak form of out-of-distribution (OoD) generalization is the ability to successfully interpolate between multiple observed distributions. One way to achieve this is through robust optimization, which seeks to minimize the worst-case risk over convex combinations of the training distributions. However, a much stronger form of OoD generalization is the ability of models to extrapolate beyond the distributions observed during training. In pursuit of strong OoD generalization, we introduce the principle of Risk Extrapolation (REx). REx can be viewed as encouraging robustness over affine combinations of training risks, by encouraging strict equality between training risks. We show conceptually how this principle enables extrapolation, and demonstrate the effectiveness and scalability of instantiations of REx on various OoD generalization tasks."]]></description>
<dc:subject>to:NB learning_theory robustness non-stationarity to_teach:childs_garden_of_statistical_learning_theory</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:c60178e00afd/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:non-stationarity"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:childs_garden_of_statistical_learning_theory"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/1908.05659">
    <title>[1908.05659] Distributionally Robust Optimization: A Review</title>
    <dc:date>2019-08-17T14:27:41+00:00</dc:date>
    <link>https://arxiv.org/abs/1908.05659</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["The concepts of risk-aversion, chance-constrained optimization, and robust optimization have developed significantly over the last decade. Statistical learning community has also witnessed a rapid theoretical and applied growth by relying on these concepts. A modeling framework, called distributionally robust optimization (DRO), has recently received significant attention in both the operations research and statistical learning communities. This paper surveys main concepts and contributions to DRO, and its relationships with robust optimization, risk-aversion, chance-constrained optimization, and function regularization."]]></description>
<dc:subject>to:NB optimization robustness probability</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:956ba641a373/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:probability"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://arxiv.org/abs/1402.6118">
    <title>[1402.6118] Approximate Models and Robust Decisions</title>
    <dc:date>2014-03-08T22:28:11+00:00</dc:date>
    <link>http://arxiv.org/abs/1402.6118</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Statistical decisions based partly or solely on predictions from probabilistic models may be sensitive to model misspecification. Statisticians are taught from an early stage that ``all models are wrong'' but little formal guidance exists on how to assess the impact of model approximation, or how to proceed when optimal actions appear sensitive to model fidelity. This article presents one potential applied framework to address this. We discuss diagnostic techniques, including graphical approaches and summary statistics, to help highlight decisions made through minimised expected loss that are sensitive to model misspecification. We then derive formal methods for decision making under model misspecification by quantifying stability of optimal actions to perturbations within a neighbourhood of model space, defined via an information (Kullback-Leibler) divergence around the approximating model. This latter approach draws heavily from recent work in the robust control, macroeconomics and financial mathematics literature. We adopt a Bayesian approach throughout although the methods are agnostic to this position."]]></description>
<dc:subject>to:NB robustness misspecification decision_theory statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:e2ba3b28fa4a/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:decision_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://arxiv.org/abs/1206.4628">
    <title>[1206.4628] Robust PCA in High-dimension: A Deterministic Approach</title>
    <dc:date>2012-06-23T13:51:59+00:00</dc:date>
    <link>http://arxiv.org/abs/1206.4628</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We consider principal component analysis for contaminated data-set in the high dimensional regime, where the dimensionality of each observation is comparable or even more than the number of observations. We propose a deterministic high-dimensional robust PCA algorithm which inherits all theoretical properties of its randomized counterpart, i.e., it is tractable, robust to contaminated points, easily kernelizable, asymptotic consistent and achieves maximal robustness -- a breakdown point of 50%. More importantly, the proposed method exhibits significantly better computational efficiency, which makes it suitable for large-scale real applications."]]></description>
<dc:subject>to:NB statistics principal_components high-dimensional_statistics robustness</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:354f23a07d65/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:principal_components"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:high-dimensional_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://jmlr.csail.mit.edu/papers/v9/debruyne08a.html">
    <title>Model Selection in Kernel Based Regression using the Influence Function</title>
    <dc:date>2012-02-05T18:55:13+00:00</dc:date>
    <link>http://jmlr.csail.mit.edu/papers/v9/debruyne08a.html</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Recent results about the robustness of kernel methods involve the analysis of influence functions. By definition the influence function is closely related to leave-one-out criteria. In statistical learning, the latter is often used to assess the generalization of a method. In statistics, the influence function is used in a similar way to analyze the statistical efficiency of a method. Links between both worlds are explored. The influence function is related to the first term of a Taylor expansion. Higher order influence functions are calculated. A recursive relation between these terms is found characterizing the full Taylor expansion. It is shown how to evaluate influence functions at a specific sample distribution to obtain an approximation of the leave-one-out error. A specific implementation is proposed using a L1 loss in the selection of the hyperparameters and a Huber loss in the estimation procedure. The parameter in the Huber loss controlling the degree of robustness is optimized as well. The resulting procedure gives good results, even when outliers are present in the data."]]></description>
<dc:subject>statistics regression kernel_estimators model_selection robustness nonparametrics cross-validation in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:466119481505/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kernel_estimators"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:model_selection"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:nonparametrics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cross-validation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism_25.html">
    <title>Abandoned Footnotes: Epistemic Arguments for Conservatism IV.5: An Addendum on Resilience</title>
    <dc:date>2010-11-25T14:20:41+00:00</dc:date>
    <link>http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism_25.html</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>conservatism institutions robustness social_life_of_the_mind</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:5fd2835c2116/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:conservatism"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:institutions"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_life_of_the_mind"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism-iv.html">
    <title>Abandoned Footnotes: Epistemic Arguments for Conservatism IV: The Resilience Argument and the “Not Dead Yet” Criterion</title>
    <dc:date>2010-11-24T23:05:31+00:00</dc:date>
    <link>http://abandonedfootnotes.blogspot.com/2010/11/epistemic-arguments-for-conservatism-iv.html</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>conservatism social_life_of_the_mind institutions robustness</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:da17002e1417/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:conservatism"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_life_of_the_mind"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:institutions"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://projecteuclid.org/DPubS?service=UI&amp;version=1.0&amp;verb=Display&amp;handle=euclid.lnms/1249305323">
    <title>Huber: On the Non-Optimality of Optimal Procedures</title>
    <dc:date>2009-12-31T18:28:36+00:00</dc:date>
    <link>http://projecteuclid.org/DPubS?service=UI&amp;version=1.0&amp;verb=Display&amp;handle=euclid.lnms/1249305323</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>statistics optimization methodological_advice robustness have_read huber.peter robust_statistics</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:327566a2d2fb/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:methodological_advice"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:huber.peter"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robust_statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://arxiv.org/abs/0804.4714">
    <title>[0804.4714] Network Structure and Dynamics, and Emergence of Robustness by</title>
    <dc:date>2008-05-02T14:38:51+00:00</dc:date>
    <link>http://arxiv.org/abs/0804.4714</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>artificial_life nk_networks evolution robustness</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:15b832e4add0/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:artificial_life"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:nk_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:evolution"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://bookstaber.com/rick/OnTheOptimalityOfCoarseBehavior.pdf">
    <title>On the Optimality of Coarse Behavior Rules (Bookstaber and Langsam, 1985)</title>
    <dc:date>2008-04-30T21:26:04+00:00</dc:date>
    <link>http://bookstaber.com/rick/OnTheOptimalityOfCoarseBehavior.pdf</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>decision-making bounded_rationality decision_theory robustness adaptive_behavior bookstaber.richard</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:f772b7988930/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:decision-making"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:bounded_rationality"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:decision_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:adaptive_behavior"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:bookstaber.richard"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://arxiv.org/abs/0710.4269">
    <title>[0710.4269] Shape, size and robustness: feasible regions in the parameter space of biochemical networks</title>
    <dc:date>2007-11-09T15:06:37+00:00</dc:date>
    <link>http://arxiv.org/abs/0710.4269</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>biochemical_networks in_NB robustness</dc:subject>
<dc:identifier>https://pinboard.in/u:cshalizi/b:362d13333440/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:biochemical_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:robustness"/>
</rdf:Bag></taxo:topics>
</item>
</rdf:RDF>