no code implementations • 22 Jan 2024 • Matteo Zecchin, Sangwoo Park, Osvaldo Simeone, Fredrik Hellström
A popular technique to achieve this goal is conformal prediction (CP), which transforms an arbitrary base predictor into a set predictor with coverage guarantees.
1 code implementation • 16 Oct 2023 • Fredrik Hellström, Benjamin Guedj
We derive generic information-theoretic and PAC-Bayesian generalization bounds involving an arbitrary convex comparator function, which measures the discrepancy between the training and population loss.
no code implementations • 8 Sep 2023 • Fredrik Hellström, Giuseppe Durisi, Benjamin Guedj, Maxim Raginsky
Over the past decades, the PAC-Bayesian approach has been established as a flexible framework to address the generalization capabilities of machine learning algorithms, and design new ones.
no code implementations • 12 Oct 2022 • Fredrik Hellström, Giuseppe Durisi
Furthermore, using the evaluated CMI, we derive a samplewise, average version of Seeger's PAC-Bayesian bound, where the convex function is the binary KL divergence.
no code implementations • 12 Oct 2022 • Fredrik Hellström, Giuseppe Durisi
Recent work has established that the conditional mutual information (CMI) framework of Steinke and Zakynthinou (2020) is expressive enough to capture generalization guarantees in terms of algorithmic stability, VC dimension, and related complexity measures for conventional learning (Harutyunyan et al., 2021, Haghifam et al., 2021).
no code implementations • 22 Oct 2020 • Fredrik Hellström, Giuseppe Durisi
If the conditional information density is bounded uniformly in the size $n$ of the training set, our bounds decay as $1/n$.
no code implementations • 28 Sep 2020 • Fredrik Hellström, Giuseppe Durisi
If the conditional information density is bounded uniformly in the size $n$ of the training set, our bounds decay as $1/n$, which is referred to as a fast rate.
no code implementations • 16 May 2020 • Fredrik Hellström, Giuseppe Durisi
We present a general approach, based on exponential inequalities, to derive bounds on the generalization error of randomized learning algorithms.
no code implementations • 20 Apr 2020 • Fredrik Hellström, Giuseppe Durisi
Our approach can be used to obtain bounds on the average generalization error as well as bounds on its tail probabilities, both for the case in which a new hypothesis is randomly generated every time the algorithm is used - as often assumed in the probably approximately correct (PAC)-Bayesian literature - and in the single-draw case, where the hypothesis is extracted only once.