Search

The raise of machine learning and deep learning led to significant improvement in several domains. This change is supported by both the dramatic rise in computation power and the collection of large datasets. Such massive datasets often include personal data which can represent a threat to privacy. Membership inference attacks are a novel direction of research which aims at recovering training data used by a learning algorithm. In this paper, we develop a mean to measure the leakage of training data leveraging a quantity appearing as a proxy of the total variation of a trained model near its training samples. We extend our work by providing a novel defense mechanism. Our contributions are supported by empirical evidence through convincing numerical experiments., Comment: Trustworthy and Socially Responsible Machine Learning (TSRML 2022) co-located with NeurIPS 2022

International audience; The use of personal data for training machine learningsystems comes with a privacy threat and measuring the levelof privacy of a model is one of the major challenges in ma-chine learning today. Identifying training data based ona trained model is a standard way of measuring the pri-vacy risks induced by the model. We develop a novel ap-proach to address the problem of membership inference inpattern recognition models, relying on information providedby adversarial examples. The strategy we propose consistsof measuring the magnitude of a perturbation necessary tobuild an adversarial example. Indeed, we argue that thisquantity reflects the likelihood of belonging to the trainingdata. Extensive numerical experiments on multivariate dataand an array of state-of-the-art target models show that ourmethod performs comparable or even outperforms state-of-the-art strategies, but without requiring any additionaltraining samples.

Understanding the complex structure of multivariate extremes is a major challenge in various fields from portfolio monitoring and environmental risk management to insurance. In the framework of multivariate Extreme Value Theory, a common characterization of extremes' dependence structure is the angular measure. It is a suitable measure to work in extreme regions as it provides meaningful insights concerning the subregions where extremes tend to concentrate their mass. The present paper develops a novel optimization-based approach to assess the dependence structure of extremes. This support identification scheme rewrites as estimating clusters of features which best capture the support of extremes. The dimension reduction technique we provide is applied to statistical learning tasks such as feature clustering and anomaly detection. Numerical experiments provide strong empirical evidence of the relevance of our approach.

The dominant approaches to text representation in natural language rely on learning embeddings on massive corpora which have convenient properties such as compositionality and distance preservation. In this paper, we develop a novel method to learn a heavy-tailed embedding with desirable regularity properties regarding the distributional tails, which allows to analyze the points far away from the distribution bulk using the framework of multivariate extreme value theory. In particular, a classifier dedicated to the tails of the proposed embedding is obtained which performance outperforms the baseline. This classifier exhibits a scale invariance property which we leverage by introducing a novel text generation method for label preserving dataset augmentation. Numerical experiments on synthetic and real text data demonstrate the relevance of the proposed framework and confirm that this method generates meaningful sentences with controllable attribute, e.g. positive or negative sentiment.