English
Related papers

Related papers: Provable Tempered Overfitting of Minimal Nets and …

200 papers

In the past decade the mathematical theory of machine learning has lagged far behind the triumphs of deep neural networks on practical challenges. However, the gap between theory and practice is gradually starting to close. In this paper I…

Machine Learning · Statistics 2021-06-01 Mikhail Belkin

We introduce a pruning algorithm that provably sparsifies the parameters of a trained model in a way that approximately preserves the model's predictive accuracy. Our algorithm uses a small batch of input points to construct a data-informed…

Machine Learning · Computer Science 2021-03-16 Cenk Baykal , Lucas Liebenwein , Igor Gilitschenski , Dan Feldman , Daniela Rus

One of the biggest issues in deep learning theory is the generalization ability of networks with huge model size. The classical learning theory suggests that overparameterized models cause overfitting. However, practically used large deep…

Machine Learning · Computer Science 2020-06-23 Taiji Suzuki , Hiroshi Abe , Tomoaki Nishimura

We develop the first (to the best of our knowledge) provably correct neural networks for a precise computational task, with the proof of correctness generated by an automated verification algorithm without any human input. Prior work on…

Machine Learning · Computer Science 2024-05-09 Rudy Bunel , Krishnamurthy Dvijotham , M. Pawan Kumar , Alessandro De Palma , Robert Stanforth

We introduce an algorithm where the individual bits representing the weights of a neural network are learned. This method allows training weights with integer values on arbitrary bit-depths and naturally uncovers sparse networks, without…

Machine Learning · Computer Science 2022-02-22 Cristian Ivan

Overparameterization and overfitting are common concerns when designing and training deep neural networks, that are often counteracted by pruning and regularization strategies. However, these strategies remain secondary to most learning…

Machine Learning · Computer Science 2020-09-01 Malena Reiners , Kathrin Klamroth , Michael Stiglmayr

Deep learning is renowned for its theory-practice gap, whereby principled theory typically fails to provide much beneficial guidance for implementation in practice. This has been highlighted recently by the benign overfitting phenomenon:…

Machine Learning · Statistics 2023-11-14 Liam Hodgkinson , Chris van der Heide , Robert Salomone , Fred Roosta , Michael W. Mahoney

Training a neural network is synonymous with learning the values of the weights. By contrast, we demonstrate that randomly weighted neural networks contain subnetworks which achieve impressive performance without ever training the weight…

Computer Vision and Pattern Recognition · Computer Science 2020-04-01 Vivek Ramanujan , Mitchell Wortsman , Aniruddha Kembhavi , Ali Farhadi , Mohammad Rastegari

In training neural networks, it is common practice to use partial gradients computed over batches, mostly very small subsets of the training set. This approach is motivated by the argument that such a partial gradient is close to the true…

Machine Learning · Computer Science 2024-11-25 Jan Spörer , Bernhard Bermeitinger , Tomas Hrycej , Niklas Limacher , Siegfried Handschuh

Recently deep neural networks have been successfully used for various classification tasks, especially for problems with massive perfectly labeled training data. However, it is often costly to have large-scale credible labels in real-world…

Machine Learning · Computer Science 2019-01-15 Mingxiao An , Yongzhou Chen , Qi Liu , Chuanren Liu , Guangyi Lv , Fangzhao Wu , Jianhui Ma

Pruning and quantization techniques have been broadly successful in reducing the number of parameters needed for large neural networks, yet theoretical justification for their empirical success falls short. We consider a randomized greedy…

Machine Learning · Computer Science 2025-12-09 Houssam El Cheairi , David Gamarnik , Rahul Mazumder

This work is substituted by the paper in arXiv:2011.14066. Stochastic gradient descent is the de facto algorithm for training deep neural networks (DNNs). Despite its popularity, it still requires fine tuning in order to achieve its best…

Machine Learning · Statistics 2020-12-02 Vatsal Shah , Anastasios Kyrillidis , Sujay Sanghavi

Deep neural networks (DNNs) defy the classical bias-variance trade-off: adding parameters to a DNN that interpolates its training data will typically improve its generalization performance. Explaining the mechanism behind this ``benign…

Machine Learning · Statistics 2023-05-02 Diego Doimo , Aldo Glielmo , Sebastian Goldt , Alessandro Laio

Deep, overparameterized regression models are notorious for their tendency to overfit. This problem is exacerbated in heteroskedastic models, which predict both mean and residual noise for each data point. At one extreme, these models fit…

Machine Learning · Statistics 2024-02-15 Eliot Wong-Toi , Alex Boyd , Vincent Fortuin , Stephan Mandt

In this paper, we study the generalization performance of min $\ell_2$-norm overfitting solutions for the neural tangent kernel (NTK) model of a two-layer neural network with ReLU activation that has no bias term. We show that, depending on…

Machine Learning · Computer Science 2023-03-09 Peizhong Ju , Xiaojun Lin , Ness B. Shroff

We propose an optimistic estimate to evaluate the best possible fitting performance of nonlinear models. It yields an optimistic sample size that quantifies the smallest possible sample size to fit/recover a target function using a…

Machine Learning · Computer Science 2023-07-19 Yaoyu Zhang , Zhongwang Zhang , Leyang Zhang , Zhiwei Bai , Tao Luo , Zhi-Qin John Xu

We rigorously evaluate three state-of-the-art techniques for inducing sparsity in deep neural networks on two large-scale learning tasks: Transformer trained on WMT 2014 English-to-German, and ResNet-50 trained on ImageNet. Across thousands…

Machine Learning · Computer Science 2019-02-27 Trevor Gale , Erich Elsen , Sara Hooker

We present a novel regularization approach to train neural networks that enjoys better generalization and test error than standard stochastic gradient descent. Our approach is based on the principles of cross-validation, where a validation…

Computer Vision and Pattern Recognition · Computer Science 2018-09-06 Simon Jenni , Paolo Favaro

We study the topmost weight matrix of neural network language models. We show that this matrix constitutes a valid word embedding. When training language models, we recommend tying the input embedding and this output embedding. We analyze…

Computation and Language · Computer Science 2017-02-22 Ofir Press , Lior Wolf

Existing generalization measures that aim to capture a model's simplicity based on parameter counts or norms fail to explain generalization in overparameterized deep neural networks. In this paper, we introduce a new, theoretically…

Machine Learning · Computer Science 2021-03-11 Lorenz Kuhn , Clare Lyle , Aidan N. Gomez , Jonas Rothfuss , Yarin Gal