English
Related papers

Related papers: Principles for Initialization and Architecture Sel…

200 papers

Initializing the weights and the biases is a key part of the training process of a neural network. Unlike the subsequent optimization phase, however, the initialization phase has gained only limited attention in the literature. In this…

Machine Learning · Computer Science 2019-09-06 Ingo Steinwart

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

Deep neural networks are typically initialized with random weights, with variances chosen to facilitate signal propagation and stable gradients. It is also believed that diversity of features is an important property of these…

Machine Learning · Computer Science 2020-07-03 Yaniv Blumenfeld , Dar Gilboa , Daniel Soudry

Graph Neural Networks (GNNs) have demonstrated remarkable performance across a spectrum of graph-related tasks, however concerns persist regarding their vulnerability to adversarial perturbations. While prevailing defense strategies focus…

Machine Learning · Computer Science 2025-10-28 Sofiane Ennadir , Johannes F. Lutzeyer , Michalis Vazirgiannis , El Houcine Bergou

Oversmoothing has been recognized as a main obstacle to building deep Graph Neural Networks (GNNs), limiting the performance. This position paper argues that the influence of oversmoothing has been overstated and advocates for a further…

Machine Learning · Computer Science 2025-06-06 MoonJeong Park , Sunghyun Choi , Jaeseung Heo , Eunhyeok Park , Dongwoo Kim

This work establishes low test error of gradient flow (GF) and stochastic gradient descent (SGD) on two-layer ReLU networks with standard initialization, in three regimes where key sets of weights rotate little (either naturally due to GF…

Machine Learning · Computer Science 2022-08-05 Matus Telgarsky

Pre-training is essential to deep learning model performance, especially in medical image analysis tasks where limited training data are available. However, existing pre-training methods are inflexible as the pre-trained weights of one…

Computer Vision and Pattern Recognition · Computer Science 2022-06-09 Fangxin Shang , Yehui Yang , Dalu Yang , Junde Wu , Xiaorong Wang , Yanwu Xu

One of the mysteries in the success of neural networks is randomly initialized first order methods like gradient descent can achieve zero training loss even though the objective function is non-convex and non-smooth. This paper demystifies…

Machine Learning · Computer Science 2019-02-06 Simon S. Du , Xiyu Zhai , Barnabas Poczos , Aarti Singh

Traditional initialisation methods, e.g. He and Xavier, have been effective in avoiding the problem of vanishing or exploding gradients in neural networks. However, they only use simple pointwise distributions, which model one-dimensional…

Computer Vision and Pattern Recognition · Computer Science 2023-10-26 Christian Harder , Moritz Fuchs , Yuri Tolkach , Anirban Mukhopadhyay

We draw connections between simple neural networks and under-determined linear systems to comprehensively explore several interesting theoretical questions in the study of neural networks. First, we emphatically show that it is unsurprising…

Numerical Analysis · Mathematics 2020-11-02 Austin R. Benson , Anil Damle , Alex Townsend

Training a high-quality deep neural network requires choosing suitable hyperparameters, which is a non-trivial and expensive process. Current works try to automatically optimize or design principles of hyperparameters, such that they can…

Machine Learning · Computer Science 2024-02-28 Wuyang Chen , Junru Wu , Zhangyang Wang , Boris Hanin

Appropriate weight initialization settings, along with the ReLU activation function, have become cornerstones of modern deep learning, enabling the training and deployment of highly effective and efficient neural network models across…

Machine Learning · Computer Science 2024-04-02 Hyunwoo Lee , Yunho Kim , Seung Yeop Yang , Hayoung Choi

In recent years significant progress has been made in successfully training recurrent neural networks (RNNs) on sequence learning problems involving long range temporal dependencies. The progress has been made on three fronts: (a)…

Neural and Evolutionary Computing · Computer Science 2016-06-24 Sachin S. Talathi , Aniket Vartak

Graph neural networks (GNNs) work remarkably well in semi-supervised node regression, yet a rigorous theory explaining when and why they succeed remains lacking. To address this gap, we study an aggregate-and-readout model that encompasses…

Machine Learning · Statistics 2026-02-20 Juntong Chen , Claire Donnat , Olga Klopp , Johannes Schmidt-Hieber

The training success, training speed and generalization ability of neural networks rely crucially on the choice of random parameter initialization. It has been shown for multiple architectures that initial dynamical isometry is particularly…

Machine Learning · Computer Science 2022-10-06 Advait Gadhikar , Rebekka Burkholz

Gradient descent (GD) type optimization methods are the standard instrument to train artificial neural networks (ANNs) with rectified linear unit (ReLU) activation. Despite the great success of GD type optimization methods in numerical…

Optimization and Control · Mathematics 2022-12-29 Arnulf Jentzen , Adrian Riekert

Activation functions are essential to introduce nonlinearity into neural networks, with the Rectified Linear Unit (ReLU) often favored for its simplicity and effectiveness. Motivated by the structural similarity between a shallow…

Machine Learning · Computer Science 2024-01-30 Jiayun Li , Yuxiao Cheng , Yiwen Lu , Zhuofan Xia , Yilin Mo , Gao Huang

Randomly initialized neural networks are known to become harder to train with increasing depth, unless architectural enhancements like residual connections and batch normalization are used. We here investigate this phenomenon by revisiting…

Machine Learning · Statistics 2020-06-15 Hadi Daneshmand , Jonas Kohler , Francis Bach , Thomas Hofmann , Aurelien Lucchi

Graph neural networks (GNNs) have become a workhorse approach for learning from data defined over irregular domains, typically by implicitly assuming that the data structure is represented by a homophilic graph. However, recent works have…

Machine Learning · Computer Science 2024-09-16 Samuel Rey , Madeline Navarro , Victor M. Tenorio , Santiago Segarra , Antonio G. Marques

LayerNorm is a critical component in modern large language models (LLMs) for stabilizing training and ensuring smooth optimization. However, it introduces significant challenges in mechanistic interpretability, outlier feature suppression,…

Machine Learning · Computer Science 2024-11-19 Nandan Kumar Jha , Brandon Reagen