English
Related papers

Related papers: A Convergence Theory Towards Practical Over-parame…

200 papers

This article concerns the expressive power of depth in neural nets with ReLU activations and bounded width. We are particularly interested in the following questions: what is the minimal width $w_{\text{min}}(d)$ so that ReLU nets of width…

Machine Learning · Statistics 2019-10-22 Boris Hanin

This survey provides an in-depth and explanatory review of the approximation properties of deep neural networks, with a focus on feed-forward and residual architectures. The primary objective is to examine how effectively neural networks…

Machine Learning · Computer Science 2024-12-18 Owen Davis , Mohammad Motamed

This work focuses on the analysis of fully connected feed forward ReLU neural networks as they approximate a given, smooth function. In contrast to conventionally studied universal approximation properties under increasing architectures,…

Machine Learning · Computer Science 2024-06-24 Erion Morina , Martin Holler

Given a training set, a loss function, and a neural network architecture, it is often taken for granted that optimal network parameters exist, and a common practice is to apply available optimization algorithms to search for them. In this…

Neural and Evolutionary Computing · Computer Science 2023-12-06 Quoc-Tung Le , Elisa Riccietti , Rémi Gribonval

We continue a long line of research aimed at proving convergence of depth 2 neural networks, trained via gradient descent, to a global minimum. Like in many previous works, our model has the following features: regression with quadratic…

Machine Learning · Computer Science 2022-12-06 Alexander Razborov

In a neural network with ReLU activations, the number of piecewise linear regions in the output can grow exponentially with depth. However, this is highly unlikely to happen when the initial parameters are sampled randomly, which therefore…

Machine Learning · Computer Science 2025-10-17 Max Milkert , David Hyde , Forrest Laine

Convex $\ell_1$ regularization using an infinite dictionary of neurons has been suggested for constructing neural networks with desired approximation guarantees, but can be affected by an arbitrary amount of over-parametrization. This can…

Optimization and Control · Mathematics 2022-06-01 Konstantin Pieper , Armenak Petrosyan

In this article, we review the literature on statistical theories of neural networks from three perspectives: approximation, training dynamics and generative models. In the first part, results on excess risks for neural networks are…

Machine Learning · Statistics 2024-09-17 Namjoon Suh , Guang Cheng

In deep learning, dense layer connectivity has become a key design principle in deep neural networks (DNNs), enabling efficient information flow and strong performance across a range of applications. In this work, we model densely connected…

Machine Learning · Computer Science 2025-10-03 Jinshu Huang , Haibin Su , Xue-Cheng Tai , Chunlin Wu

We propose and analyze a new family of algorithms for training neural networks with ReLU activations. Our algorithms are based on the technique of alternating minimization: estimating the activation patterns of each ReLU for all given…

Machine Learning · Computer Science 2018-10-12 Gauri Jagatap , Chinmay Hegde

Training deep neural networks is a challenging non-convex optimization problem. Recent work has proven that the strong duality holds (which means zero duality gap) for regularized finite-width two-layer ReLU networks and consequently…

Machine Learning · Computer Science 2023-03-08 Yifei Wang , Tolga Ergen , Mert Pilanci

While neural networks are used for classification tasks across domains, a long-standing open problem in machine learning is determining whether neural networks trained using standard procedures are optimal for classification, i.e., whether…

Machine Learning · Computer Science 2023-05-03 Adityanarayanan Radhakrishnan , Mikhail Belkin , Caroline Uhler

It has been shown that gradient descent can yield the zero training loss in the over-parametrized regime (the width of the neural networks is much larger than the number of data points). In this work, combining the ideas of some existing…

Optimization and Control · Mathematics 2019-11-05 Lei Li

A new network with super approximation power is introduced. This network is built with Floor ($\lfloor x\rfloor$) or ReLU ($\max\{0,x\}$) activation function in each neuron and hence we call such networks Floor-ReLU networks. For any…

Machine Learning · Computer Science 2021-03-30 Zuowei Shen , Haizhao Yang , Shijun Zhang

Training deep neural networks with stochastic gradient descent (SGD) can often achieve zero training loss on real-world tasks although the optimization landscape is known to be highly non-convex. To understand the success of SGD for…

Machine Learning · Statistics 2020-06-15 Yiping Lu , Chao Ma , Yulong Lu , Jianfeng Lu , Lexing Ying

Deep learning models have gained great success in many real-world applications. However, most existing networks are typically designed in heuristic manners, thus lack of rigorous mathematical principles and derivations. Several recent…

Computer Vision and Pattern Recognition · Computer Science 2017-12-18 Risheng Liu , Xin Fan , Shichao Cheng , Xiangyu Wang , Zhongxuan Luo

We study training one-hidden-layer ReLU networks in the neural tangent kernel (NTK) regime, where the networks' biases are initialized to some constant rather than zero. We prove that under such initialization, the neural network will have…

Machine Learning · Computer Science 2024-10-31 Hongru Yang , Ziyu Jiang , Ruizhe Zhang , Yingbin Liang , Zhangyang Wang

In practice, multi-task learning (through learning features shared among tasks) is an essential property of deep neural networks (NNs). While infinite-width limits of NNs can provide good intuition for their generalization behavior, the…

Machine Learning · Computer Science 2022-10-21 Jakob Heiss , Josef Teichmann , Hanna Wutte

We develop exact representations of training two-layer neural networks with rectified linear units (ReLUs) in terms of a single convex program with number of variables polynomial in the number of training samples and the number of hidden…

Machine Learning · Computer Science 2020-08-18 Mert Pilanci , Tolga Ergen

Training large and highly accurate deep learning (DL) models is computationally costly. This cost is in great part due to the excessive number of trained parameters, which are well-known to be redundant and compressible for the execution…

Machine Learning · Computer Science 2019-04-11 Mojan Javaheripi , Bita Darvish Rouhani , Farinaz Koushanfar