Claudio Gentile
YOU?
Author Swipe
View article: Faster margin maximization rates for generic and adversarially robust optimization methods
Faster margin maximization rates for generic and adversarially robust optimization methods Open
First-order optimization methods tend to inherently favor certain solutions over others when minimizing an underdetermined training objective that has multiple global optima. This phenomenon, known as implicit bias , plays a critical role …
View article: Nearly Optimal Sample Complexity for Learning with Label Proportions
Nearly Optimal Sample Complexity for Learning with Label Proportions Open
We investigate Learning from Label Proportions (LLP), a partial information setting where examples in a training set are grouped into bags, and only aggregate label values in each bag are available. Despite the partial observability, the g…
View article: Auditing Privacy Mechanisms via Label Inference Attacks
Auditing Privacy Mechanisms via Label Inference Attacks Open
We propose reconstruction advantage measures to audit label privatization mechanisms. A reconstruction advantage measure quantifies the increase in an attacker's ability to infer the true label of an unlabeled example when provided with a …
View article: Fast and Effective GNN Training through Sequences of Random Path Graphs
Fast and Effective GNN Training through Sequences of Random Path Graphs Open
We present GERN, a novel scalable framework for training GNNs in node classification tasks, based on effective resistance, a standard tool in spectral graph theory. Our method progressively refines the GNN weights on a sequence of random s…
View article: Data-Driven Online Model Selection With Regret Guarantees
Data-Driven Online Model Selection With Regret Guarantees Open
We consider model selection for sequential decision making in stochastic environments with bandit feedback, where a meta-learner has at its disposal a pool of base learners, and decides on the fly which action to take based on the policies…
View article: Leveraging User-Triggered Supervision in Contextual Bandits
Leveraging User-Triggered Supervision in Contextual Bandits Open
We study contextual bandit (CB) problems, where the user can sometimes respond with the best action in a given context. Such an interaction arises, for example, in text prediction or autocompletion settings, where a poor suggestion is simp…
View article: Easy Learning from Label Proportions
Easy Learning from Label Proportions Open
We consider the problem of Learning from Label Proportions (LLP), a weakly supervised classification setup where instances are grouped into "bags", and only the frequency of class labels at each bag is available. Albeit, the objective of t…
View article: A Contextual Bandit Approach for Learning to Plan in Environments with Probabilistic Goal Configurations
A Contextual Bandit Approach for Learning to Plan in Environments with Probabilistic Goal Configurations Open
Object-goal navigation (Object-nav) entails searching, recognizing and navigating to a target object. Object-nav has been extensively studied by the Embodied-AI community, but most solutions are often restricted to considering static objec…
View article: Le nuove Norme sui delitti riservati alla Congregazione per la Dottrina della Fede. Prime considerazioni
Le nuove Norme sui delitti riservati alla Congregazione per la Dottrina della Fede. Prime considerazioni Open
SOMMARIO: 1. Inquadramento giuridico - 2. La nuova edizione delle Norme - 3. Conclusioni.
The new Norms Regarding Delicts Reserved To The Congregation For The Doctrine Of The Faith. First considerations
ABSTRACT: On December 7, 2021, the…
View article: Fast Rates in Pool-Based Batch Active Learning
Fast Rates in Pool-Based Batch Active Learning Open
We consider a batch active learning scenario where the learner adaptively issues batches of points to a labeling oracle. Sampling labels in batches is highly desirable in practice due to the smaller number of interactive rounds with the la…
View article: Batch Active Learning at Scale
Batch Active Learning at Scale Open
The ability to train complex and highly effective models often requires an abundance of training data, which can easily become a bottleneck in cost, time, and computational resources. Batch active learning, which adaptively issues batched …
View article: On Learning to Rank Long Sequences with Contextual Bandits
On Learning to Rank Long Sequences with Contextual Bandits Open
Motivated by problems of learning to rank long item sequences, we introduce a variant of the cascading bandit model that considers flexible length sequences with varying rewards and losses. We formulate two generative models for this probl…
View article: Neural Active Learning with Performance Guarantees
Neural Active Learning with Performance Guarantees Open
We investigate the problem of active learning in the streaming setting in non-parametric regimes, where the labels are stochastically generated from a class of functions on which we make no assumptions whatsoever. We rely on recently propo…
View article: Regret Bound Balancing and Elimination for Model Selection in Bandits and RL
Regret Bound Balancing and Elimination for Model Selection in Bandits and RL Open
We propose a simple model selection approach for algorithms in stochastic bandit and reinforcement learning problems. As opposed to prior work that (implicitly) assumes knowledge of the optimal regret, we only require that each base algori…
View article: Adaptive Region-Based Active Learning
Adaptive Region-Based Active Learning Open
We present a new active learning algorithm that adaptively partitions the input space into a finite number of regions, and subsequently seeks a distinct predictor for each region, both phases actively requesting labels. We prove theoretica…
View article: On Similarity Prediction and Pairwise Clustering
On Similarity Prediction and Pairwise Clustering Open
We consider the problem of clustering a finite set of items from pairwise similarity information. Unlike what is done in the literature on this subject, we do so in a passive learning setting, and with no specific constraints on the cluste…
View article: Identification of Mobile Phones Using the Built-In Magnetometers Stimulated by Motion Patterns
Identification of Mobile Phones Using the Built-In Magnetometers Stimulated by Motion Patterns Open
We investigate the identification of mobile phones through their built-in magnetometers. These electronic components have started to be widely deployed in mass market phones in recent years, and they can be exploited to uniquely identify m…
View article: Online Learning with Abstention
Online Learning with Abstention Open
We present an extensive study of the key problem of online learning where algorithms are allowed to abstain from making predictions. In the adversarial setting, we show how existing online algorithms and guarantees can be adapted to this p…
View article: Online Nonparametric Learning, Chaining, and the Role of Partial Feedback
Online Nonparametric Learning, Chaining, and the Role of Partial Feedback Open
We investigate contextual online learning with nonparametric (Lipschitz) comparison classes under different assumptions on losses and feedback information. For full information feedback and Lipschitz losses, we characterize the minimax reg…
View article: Nonstochastic Multi-Armed Bandits with Graph-Structured Feedback
Nonstochastic Multi-Armed Bandits with Graph-Structured Feedback Open
We present and study a partial-information model of online learning, where a decision maker repeatedly chooses from a finite set of actions, and observes some subset of the associated losses. This naturally models several situations where …
View article: Nonstochastic Multi-Armed Bandits with Graph-Structured Feedback
Nonstochastic Multi-Armed Bandits with Graph-Structured Feedback Open
We introduce and study a partial-information model of online learning, where a decision maker repeatedly chooses from a finite set of actions and observes some subset of the associated losses. This setting naturally models several situatio…
View article: Online Context-Dependent Clustering in Recommendations based on Exploration-Exploitation Algorithms
Online Context-Dependent Clustering in Recommendations based on Exploration-Exploitation Algorithms Open
We investigate two context-dependent clustering techniques for content recommendation based on exploration-exploitation strategies in contextual multi-armed bandit settings. Our algorithms dynamically group users based on the items under c…
View article: On Context-Dependent Clustering of Bandits
On Context-Dependent Clustering of Bandits Open
We investigate a novel cluster-of-bandit algorithm CAB for collaborative recommendation tasks that implements the underlying feedback sharing mechanism by estimating the neighborhood of users in a context-dependent manner. CAB makes sharp …
View article: On the Troll-Trust Model for Edge Sign Prediction in Social Networks
On the Troll-Trust Model for Edge Sign Prediction in Social Networks Open
In the problem of edge sign prediction, we are given a directed graph (representing a social network), and our task is to predict the binary labels of the edges (i.e., the positive or negative nature of the social relationships). Many succ…
View article: Graph Clustering Bandits for Recommendation
Graph Clustering Bandits for Recommendation Open
We investigate an efficient context-dependent clustering technique for recommender systems based on exploration-exploitation strategies through multi-armed bandits over multiple users. Our algorithm dynamically groups users based on their …
View article: Delay and Cooperation in Nonstochastic Bandits
Delay and Cooperation in Nonstochastic Bandits Open
We study networks of communicating learning agents that cooperate to solve a common nonstochastic bandit problem. Agents use an underlying communication network to get messages about actions selected by other agents, and drop messages that…
View article: Collaborative Filtering Bandits
Collaborative Filtering Bandits Open
Classical collaborative filtering, and content-based filtering methods try to learn a static recommendation model given training data. These approaches are far from ideal in highly dynamic recommendation domains such as news recommendation…