Randomization matters How to defend against strong adversarial attacks
Jul 12, 2020
Speakers
Rafael Pinot
Speaker · 0 followers
Raphael Ettedgui
Speaker · 0 followers
Geovani Rizk
Speaker · 0 followers
Yann Chevaleyre
Speaker · 0 followers
Jamal Atif
Speaker · 0 followers
About
Is there a classifier that ensures optimal robustness against all adversarial attacks? This paper answers this question by adopting a game-theoretic point of view. We show that adversarial attacks and defenses form an infinite zero-sum game where classical results (e.g. Nash or Sion theorems) do not apply. We demonstrate the non-existence of a Nash equilibrium in our game when the classifier and the adversary are both deterministic, hence giving a negative answer to the above question in the deterministic regime. Nonetheless, the question remains open in the randomized regime. We tackle this problem by showing that, under mild conditions on the dataset distribution, any deterministic classifier can be outperformed by a randomized one. This gives arguments for using randomization, and leads us to a new algorithm for building randomized classifiers that are robust to strong adversarial attacks. Empirical results validate our theoretical analysis, and show that our defense method considerably outperforms Adversarial Training against state-of-the-art attacks.Is there a classifier that ensures optimal robustness against all adversarial attacks? This paper answers this question by adopting a game-theoretic point of view. We show that adversarial attacks and defenses form an infinite zero-sum game where classical results (e.g. Nash or Sion theorems) do not apply. We demonstrate the non-existence of a Nash equilibrium in our game when the classifier and the adversary are both deterministic, hence giving a negative answer to the above question in the det…
Categories
AI & Data Science
Category · 10.8k presentations
Cybersecurity
Category · 59 presentations
About ICML 2020
The International Conference on Machine Learning (ICML) is the premier gathering of professionals dedicated to the advancement of the branch of artificial intelligence known as machine learning. ICML is globally renowned for presenting and publishing cutting-edge research on all aspects of machine learning used in closely related areas like artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, and robotics. ICML is one of the fastest growing artificial intelligence conferences in the world. Participants at ICML span a wide range of backgrounds, from academic and industrial researchers, to entrepreneurs and engineers, to graduate students and postdocs.
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
An Explicitly Relational Neural Network Architecture
Watch later
On a Projective Ensemble Approach to Two Sample Test for Equality of Distributions
Watch later
Zhimei Li, …
Kernel interpolation with continuous volume sampling
Watch later
Online metric algorithms with untrusted predictions
Watch later
Do We Need Zero Training Loss After Achieving Zero Training Error?
Watch later
Boosting Deep Neural Network Efficiency with Dual-Module Inference
Watch later
Liu Liu, …