English
Related papers

Related papers: Landscape Complexity for the Empirical Risk of Gen…

200 papers

We present a method to obtain the average and the typical value of the number of critical points of the empirical risk landscape for generalized linear estimation problems and variants. This represents a substantial extension of previous…

Machine Learning · Statistics 2023-01-19 Antoine Maillard , Gérard Ben Arous , Giulio Biroli

We consider the landscape of empirical risk minimization for high-dimensional Gaussian single-index models (generalized linear models). The objective is to recover an unknown signal $\boldsymbol{\theta}^\star \in \mathbb{R}^d$ (where $d \gg…

Machine Learning · Statistics 2026-02-23 Antoine Maillard , Tony Bonnaire , Giulio Biroli

We study rough high-dimensional landscapes in which an increasingly stronger preference for a given configuration emerges. Such energy landscapes arise in glass physics and inference. In particular we focus on random Gaussian functions, and…

Disordered Systems and Neural Networks · Physics 2019-01-09 Valentina Ros , Gerard Ben Arous , Giulio Biroli , Chiara Cammarota

We present a null model for single- and multi-layered complex systems constructed using homogeneous and isotropic random Gaussian maps. By means of a Kac-Rice formalism, we show that the mean number of fixed points can be calculated as the…

Mathematical Physics · Physics 2018-11-14 J. R. Ipsen , P. J. Forrester

Most high-dimensional estimation and prediction methods propose to minimize a cost function (empirical risk) that is written as a sum of losses associated to each data point. In this paper we focus on the case of non-convex losses, which is…

Machine Learning · Statistics 2017-01-17 Song Mei , Yu Bai , Andrea Montanari

We consider a general model for high-dimensional empirical risk minimization whereby the data $\mathbf{x}_i$ are $d$-dimensional Gaussian vectors, the model is parametrized by $\mathbf{\Theta}\in\mathbb{R}^{d\times k}$, and the loss depends…

Machine Learning · Statistics 2026-01-26 Kiana Asgari , Andrea Montanari , Basil Saeed

In this article we propose a general class of risk measures which can be used for data based evaluation of parametric models. The loss function is defined as generalized quadratic distance between the true density and the proposed model.…

Statistics Theory · Mathematics 2007-10-02 Surajit Ray , Bruce G. Lindsay

Significant advances have been made recently on training neural networks, where the main challenge is in solving an optimization problem with abundant critical points. However, existing approaches to address this issue crucially rely on a…

Machine Learning · Computer Science 2019-02-28 Weihao Gao , Ashok Vardhan Makkuva , Sewoong Oh , Pramod Viswanath

Motivated by current interest in understanding statistical properties of random landscapes in high-dimensional spaces, we consider a model of the landscape in $\mathbb{R}^N$ obtained by superimposing $M>N$ plane waves of random wavevectors…

Statistical Mechanics · Physics 2022-09-14 Bertrand Lacroix-A-Chez-Toine , Sirio Belga Fedeli , Yan V. Fyodorov

This paper develops several average-case reduction techniques to show new hardness results for three central high-dimensional statistics problems, implying a statistical-computational gap induced by robustness, a detection-recovery gap and…

Computational Complexity · Computer Science 2020-05-20 Matthew Brennan , Guy Bresler

This paper characterizes the annealed, topological complexity (both of total critical points and of local minima) of the elastic manifold. This classical model of a disordered elastic system captures point configurations with…

Probability · Mathematics 2021-05-12 Gérard Ben Arous , Paul Bourgade , Benjamin McKenna

Viewing neural network models in terms of their loss landscapes has a long history in the statistical mechanics approach to learning, and in recent years it has received attention within machine learning proper. Among other things, local…

Machine Learning · Computer Science 2021-12-14 Yaoqing Yang , Liam Hodgkinson , Ryan Theisen , Joe Zou , Joseph E. Gonzalez , Kannan Ramchandran , Michael W. Mahoney

Complexity measures are essential to understand complex systems and there are numerous definitions to analyze one-dimensional data. However, extensions of these approaches to two or higher-dimensional data, such as images, are much less…

Data Analysis, Statistics and Probability · Physics 2012-12-27 H. V. Ribeiro , L. Zunino , E. K. Lenzi , P. A. Santoro , R. S. Mendes

Recent work has established clear links between the generalization performance of trained neural networks and the geometry of their loss landscape near the local minima to which they converge. This suggests that qualitative and quantitative…

Machine Learning · Computer Science 2022-01-28 Stefan Horoi , Jessie Huang , Bastian Rieck , Guillaume Lajoie , Guy Wolf , Smita Krishnaswamy

In this paper, we propose a method to perform empirical analysis of the loss landscape of machine learning (ML) models. The method is applied to two ML models for scientific sensing, which necessitates quantization to be deployed and are…

Machine Learning · Computer Science 2025-02-17 Tommaso Baldi , Javier Campos , Olivia Weng , Caleb Geniesse , Nhan Tran , Ryan Kastner , Alessandro Biondi

We consider the Generalized Lotka-Volterra system of equations with all-to-all, random asymmetric interactions describing high-dimensional, very diverse and well-mixed ecosystems. We analyze the multiple equilibria phase of the model and…

Disordered Systems and Neural Networks · Physics 2023-06-27 Valentina Ros , Felix Roy , Giulio Biroli , Guy Bunin

This paper analyzes the convergence and generalization of training a one-hidden-layer neural network when the input features follow the Gaussian mixture model consisting of a finite number of Gaussian distributions. Assuming the labels are…

Machine Learning · Computer Science 2023-01-30 Hongkang Li , Shuai Zhang , Meng Wang

The clear understanding of the non-convex landscape of neural network is a complex incomplete problem. This paper studies the landscape of linear (residual) network, the simplified version of the nonlinear network. By treating the gradient…

Algebraic Geometry · Mathematics 2021-02-09 Xiuyi Yang

This paper tackles the problem of robust covariance matrix estimation when the data is incomplete. Classical statistical estimation methodologies are usually built upon the Gaussian assumption, whereas existing robust estimation ones assume…

We consider the problem of learning a one-hidden-layer neural network: we assume the input $x\in \mathbb{R}^d$ is from Gaussian distribution and the label $y = a^\top \sigma(Bx) + \xi$, where $a$ is a nonnegative vector in $\mathbb{R}^m$…

Machine Learning · Computer Science 2017-11-06 Rong Ge , Jason D. Lee , Tengyu Ma
‹ Prev 1 2 3 10 Next ›