Density of states in neural networks: an in-depth exploration of learning in parameter space
Learning in neural networks critically hinges on the intricate geometry of the loss landscape
associated with a given task. Traditionally, most research has focused on finding specific …
associated with a given task. Traditionally, most research has focused on finding specific …
Building Conformal Prediction Intervals with Approximate Message Passing
L Clarté, L Zdeborová - arXiv preprint arXiv:2410.16493, 2024 - arxiv.org
Conformal prediction has emerged as a powerful tool for building prediction intervals that
are valid in a distribution-free way. However, its evaluation may be computationally costly …
are valid in a distribution-free way. However, its evaluation may be computationally costly …
Phase transition and higher order analysis of Lq regularization under dependence
We study the problem of estimating a-sparse signal from a set of noisy observations under
the model, where is the measurement matrix the row of which is drawn from distribution. We …
the model, where is the measurement matrix the row of which is drawn from distribution. We …
[PDF][PDF] Analysis of Bootstrap and Subsampling in High-dimensional Regularized Regression
LCA Vandenbroucque, G Dalle, BLFK Lenka - raw.githubusercontent.com
We investigate popular resampling methods for estimating the uncertainty of statistical
models, such as subsampling, bootstrap and the jackknife, and their performance in high …
models, such as subsampling, bootstrap and the jackknife, and their performance in high …