A Boundary Tilting Persepective on the Phenomenon of Adversarial Examples

by   Thomas Tanay, et al.

Deep neural networks have been shown to suffer from a surprising weakness: their classification outputs can be changed by small, non-random perturbations of their inputs. This adversarial example phenomenon has been explained as originating from deep networks being "too linear" (Goodfellow et al., 2014). We show here that the linear explanation of adversarial examples presents a number of limitations: the formal argument is not convincing, linear classifiers do not always suffer from the phenomenon, and when they do their adversarial examples are different from the ones affecting deep networks. We propose a new perspective on the phenomenon. We argue that adversarial examples exist when the classification boundary lies close to the submanifold of sampled data, and present a mathematical analysis of this new perspective in the linear case. We define the notion of adversarial strength and show that it can be reduced to the deviation angle between the classifier considered and the nearest centroid classifier. Then, we show that the adversarial strength can be made arbitrarily high independently of the classification performance due to a mechanism that we call boundary tilting. This result leads us to defining a new taxonomy of adversarial examples. Finally, we show that the adversarial strength observed in practice is directly dependent on the level of regularisation used and the strongest adversarial examples, symptomatic of overfitting, can be avoided by using a proper level of regularisation.


page 2

page 4

page 5

page 14

page 16


Explaining and Harnessing Adversarial Examples

Several machine learning models, including neural networks, consistently...

zoNNscan : a boundary-entropy index for zone inspection of neural models

The training of deep neural network classifiers results in decision boun...

Adversarial Examples in Multi-Layer Random ReLU Networks

We consider the phenomenon of adversarial examples in ReLU networks with...

Benign Overfitting in Adversarially Robust Linear Classification

"Benign overfitting", where classifiers memorize noisy training data yet...

When adversarial examples are excusable

Neural networks work remarkably well in practice and theoretically they ...

An Analytic Framework for Robust Training of Artificial Neural Networks

The reliability of a learning model is key to the successful deployment ...

Adversarial Examples Target Topological Holes in Deep Networks

It is currently unclear why adversarial examples are easy to construct f...

Please sign up or login with your details

Forgot password? Click here to reset