Sharp Statistical Guarantees for Adversarially Robust Gaussian
Classification
- URL: http://arxiv.org/abs/2006.16384v1
- Date: Mon, 29 Jun 2020 21:06:52 GMT
- Title: Sharp Statistical Guarantees for Adversarially Robust Gaussian
Classification
- Authors: Chen Dan, Yuting Wei, Pradeep Ravikumar
- Abstract summary: We provide the first result of the optimal minimax guarantees for the excess risk for adversarially robust classification.
Results are stated in terms of the Adversarial Signal-to-Noise Ratio (AdvSNR), which generalizes a similar notion for standard linear classification to the adversarial setting.
- Score: 54.22421582955454
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Adversarial robustness has become a fundamental requirement in modern machine
learning applications. Yet, there has been surprisingly little statistical
understanding so far. In this paper, we provide the first result of the optimal
minimax guarantees for the excess risk for adversarially robust classification,
under Gaussian mixture model proposed by \cite{schmidt2018adversarially}. The
results are stated in terms of the Adversarial Signal-to-Noise Ratio (AdvSNR),
which generalizes a similar notion for standard linear classification to the
adversarial setting. For the Gaussian mixtures with AdvSNR value of $r$, we
establish an excess risk lower bound of order $\Theta(e^{-(\frac{1}{8}+o(1))
r^2} \frac{d}{n})$ and design a computationally efficient estimator that
achieves this optimal rate. Our results built upon minimal set of assumptions
while cover a wide spectrum of adversarial perturbations including $\ell_p$
balls for any $p \ge 1$.
Related papers
- Robust Linear Predictions: Analyses of Uniform Concentration, Fast Rates
and Model Misspecification [16.0817847880416]
We offer a unified framework that includes a broad variety of linear prediction problems on a Hilbert space.
We show that for misspecification level $epsilon$, these estimators achieve an error rate of $O(maxleft|mathcalO|1/2n-1/2, |mathcalI|1/2n-1 right+epsilon)$, matching the best-known rates in literature.
arXiv Detail & Related papers (2022-01-06T08:51:08Z) - Consistent Estimation for PCA and Sparse Regression with Oblivious
Outliers [13.244654316770815]
We develop machinery to design efficiently computable and consistent estimators.
For sparse regression, we achieve consistency for optimal sample size $ngsim (klog d)/alpha2$.
In the context of PCA, we attain optimal error guarantees under broad spikiness assumptions on the parameter matrix.
arXiv Detail & Related papers (2021-11-04T15:59:44Z) - On the robustness of randomized classifiers to adversarial examples [11.359085303200981]
We introduce a new notion of robustness for randomized classifiers, enforcing local Lipschitzness using probability metrics.
We show that our results are applicable to a wide range of machine learning models under mild hypotheses.
All robust models we trained models can simultaneously achieve state-of-the-art accuracy.
arXiv Detail & Related papers (2021-02-22T10:16:58Z) - Large-Scale Methods for Distributionally Robust Optimization [53.98643772533416]
We prove that our algorithms require a number of evaluations gradient independent of training set size and number of parameters.
Experiments on MNIST and ImageNet confirm the theoretical scaling of our algorithms, which are 9--36 times more efficient than full-batch methods.
arXiv Detail & Related papers (2020-10-12T17:41:44Z) - Optimal Robust Linear Regression in Nearly Linear Time [97.11565882347772]
We study the problem of high-dimensional robust linear regression where a learner is given access to $n$ samples from the generative model $Y = langle X,w* rangle + epsilon$
We propose estimators for this problem under two settings: (i) $X$ is L4-L2 hypercontractive, $mathbbE [XXtop]$ has bounded condition number and $epsilon$ has bounded variance and (ii) $X$ is sub-Gaussian with identity second moment and $epsilon$ is
arXiv Detail & Related papers (2020-07-16T06:44:44Z) - Estimating Principal Components under Adversarial Perturbations [25.778123431786653]
We study a natural model of robustness for high-dimensional statistical estimation problems.
Our model is motivated by emerging paradigms such as low precision machine learning and adversarial training.
arXiv Detail & Related papers (2020-05-31T20:27:19Z) - Towards Assessment of Randomized Smoothing Mechanisms for Certifying
Adversarial Robustness [50.96431444396752]
We argue that the main difficulty is how to assess the appropriateness of each randomized mechanism.
We first conclude that the Gaussian mechanism is indeed an appropriate option to certify $ell$-norm.
Surprisingly, we show that the Gaussian mechanism is also an appropriate option for certifying $ell_infty$-norm, instead of the Exponential mechanism.
arXiv Detail & Related papers (2020-05-15T03:54:53Z) - Toward Adversarial Robustness via Semi-supervised Robust Training [93.36310070269643]
Adrial examples have been shown to be the severe threat to deep neural networks (DNNs)
We propose a novel defense method, the robust training (RT), by jointly minimizing two separated risks ($R_stand$ and $R_rob$)
arXiv Detail & Related papers (2020-03-16T02:14:08Z) - Curse of Dimensionality on Randomized Smoothing for Certifiable
Robustness [151.67113334248464]
We show that extending the smoothing technique to defend against other attack models can be challenging.
We present experimental results on CIFAR to validate our theory.
arXiv Detail & Related papers (2020-02-08T22:02:14Z) - Extrapolation Towards Imaginary $0$-Nearest Neighbour and Its Improved
Convergence Rate [13.985534521589257]
We propose a novel multiscale $k$-NN (MS-$k$-NN) that extrapolates unweighted $k$-NN estimators from several $k ge 1$ values to $k=0$.
Our method implicitly computes optimal real-valued weights that are adaptive to the query and its neighbour points.
arXiv Detail & Related papers (2020-02-08T00:32:12Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.