Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Zhenyu Liao

Zhenyu Liao

Hessian Eigenspectra of More Realistic Nonlinear Models


Mar 17, 2021
Zhenyu Liao, Michael W. Mahoney

Figure 1 for Hessian Eigenspectra of More Realistic Nonlinear Models
Figure 2 for Hessian Eigenspectra of More Realistic Nonlinear Models
Figure 3 for Hessian Eigenspectra of More Realistic Nonlinear Models
Figure 4 for Hessian Eigenspectra of More Realistic Nonlinear Models
* Identical to v1, except for the inclusion of some additional references 
  Access Paper or Ask Questions

Sparse sketches with small inversion bias


Nov 21, 2020
MichaƂ DereziƄski, Zhenyu Liao, Edgar Dobriban, Michael W. Mahoney


  Access Paper or Ask Questions

Kernel regression in high dimension: Refined analysis beyond double descent


Oct 06, 2020
Fanghui Liu, Zhenyu Liao, Johan A. K. Suykens

Figure 1 for Kernel regression in high dimension: Refined analysis beyond double descent
Figure 2 for Kernel regression in high dimension: Refined analysis beyond double descent
Figure 3 for Kernel regression in high dimension: Refined analysis beyond double descent
Figure 4 for Kernel regression in high dimension: Refined analysis beyond double descent
* 30 pages, 13 figures 
  Access Paper or Ask Questions

Sparse Quantized Spectral Clustering


Oct 03, 2020
Zhenyu Liao, Romain Couillet, Michael W. Mahoney

Figure 1 for Sparse Quantized Spectral Clustering
Figure 2 for Sparse Quantized Spectral Clustering
Figure 3 for Sparse Quantized Spectral Clustering
Figure 4 for Sparse Quantized Spectral Clustering

  Access Paper or Ask Questions

Precise expressions for random projections: Low-rank approximation and randomized Newton


Jun 18, 2020
MichaƂ DereziƄski, Feynman Liang, Zhenyu Liao, Michael W. Mahoney

Figure 1 for Precise expressions for random projections: Low-rank approximation and randomized Newton
Figure 2 for Precise expressions for random projections: Low-rank approximation and randomized Newton
Figure 3 for Precise expressions for random projections: Low-rank approximation and randomized Newton
Figure 4 for Precise expressions for random projections: Low-rank approximation and randomized Newton

  Access Paper or Ask Questions

A random matrix analysis of random Fourier features: beyond the Gaussian kernel, a precise phase transition, and the corresponding double descent


Jun 09, 2020
Zhenyu Liao, Romain Couillet, Michael W. Mahoney

Figure 1 for A random matrix analysis of random Fourier features: beyond the Gaussian kernel, a precise phase transition, and the corresponding double descent
Figure 2 for A random matrix analysis of random Fourier features: beyond the Gaussian kernel, a precise phase transition, and the corresponding double descent
Figure 3 for A random matrix analysis of random Fourier features: beyond the Gaussian kernel, a precise phase transition, and the corresponding double descent
Figure 4 for A random matrix analysis of random Fourier features: beyond the Gaussian kernel, a precise phase transition, and the corresponding double descent

  Access Paper or Ask Questions

Towards Efficient Training for Neural Network Quantization


Dec 21, 2019
Qing Jin, Linjie Yang, Zhenyu Liao

Figure 1 for Towards Efficient Training for Neural Network Quantization
Figure 2 for Towards Efficient Training for Neural Network Quantization
Figure 3 for Towards Efficient Training for Neural Network Quantization
Figure 4 for Towards Efficient Training for Neural Network Quantization
* 23 pages, 8 figures 
  Access Paper or Ask Questions

AdaBits: Neural Network Quantization with Adaptive Bit-Widths


Dec 20, 2019
Qing Jin, Linjie Yang, Zhenyu Liao

Figure 1 for AdaBits: Neural Network Quantization with Adaptive Bit-Widths
Figure 2 for AdaBits: Neural Network Quantization with Adaptive Bit-Widths
Figure 3 for AdaBits: Neural Network Quantization with Adaptive Bit-Widths
Figure 4 for AdaBits: Neural Network Quantization with Adaptive Bit-Widths
* 10 pages, 6 figures, submitted to CVPR 2020 
  Access Paper or Ask Questions

Inner-product Kernels are Asymptotically Equivalent to Binary Discrete Kernels


Sep 15, 2019
Zhenyu Liao, Romain Couillet

Figure 1 for Inner-product Kernels are Asymptotically Equivalent to Binary Discrete Kernels
Figure 2 for Inner-product Kernels are Asymptotically Equivalent to Binary Discrete Kernels
Figure 3 for Inner-product Kernels are Asymptotically Equivalent to Binary Discrete Kernels
Figure 4 for Inner-product Kernels are Asymptotically Equivalent to Binary Discrete Kernels

  Access Paper or Ask Questions

Complete Dictionary Learning via $\ell^4$-Norm Maximization over the Orthogonal Group


Jul 10, 2019
Yuexiang Zhai, Zitong Yang, Zhenyu Liao, John Wright, Yi Ma

Figure 1 for Complete Dictionary Learning via $\ell^4$-Norm Maximization over the Orthogonal Group
Figure 2 for Complete Dictionary Learning via $\ell^4$-Norm Maximization over the Orthogonal Group
Figure 3 for Complete Dictionary Learning via $\ell^4$-Norm Maximization over the Orthogonal Group
Figure 4 for Complete Dictionary Learning via $\ell^4$-Norm Maximization over the Orthogonal Group

  Access Paper or Ask Questions

High Dimensional Classification via Empirical Risk Minimization: Improvements and Optimality


May 31, 2019
Xiaoyi Mai, Zhenyu Liao

Figure 1 for High Dimensional Classification via Empirical Risk Minimization: Improvements and Optimality
Figure 2 for High Dimensional Classification via Empirical Risk Minimization: Improvements and Optimality
Figure 3 for High Dimensional Classification via Empirical Risk Minimization: Improvements and Optimality
Figure 4 for High Dimensional Classification via Empirical Risk Minimization: Improvements and Optimality

  Access Paper or Ask Questions

Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses


Apr 01, 2019
Yingwei Li, Song Bai, Cihang Xie, Zhenyu Liao, Xiaohui Shen, Alan L. Yuille

Figure 1 for Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses
Figure 2 for Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses
Figure 3 for Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses
Figure 4 for Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses
* The code is available here: https://github.com/LiYingwei/Regional-Homogeneity 
  Access Paper or Ask Questions

A Geometric Approach of Gradient Descent Algorithms in Neural Networks


Nov 08, 2018
Yacine Chitour, Zhenyu Liao, Romain Couillet

Figure 1 for A Geometric Approach of Gradient Descent Algorithms in Neural Networks
Figure 2 for A Geometric Approach of Gradient Descent Algorithms in Neural Networks
Figure 3 for A Geometric Approach of Gradient Descent Algorithms in Neural Networks
Figure 4 for A Geometric Approach of Gradient Descent Algorithms in Neural Networks
* Preprint. Work in progress 
  Access Paper or Ask Questions

On the Spectrum of Random Features Maps of High Dimensional Data


Jul 20, 2018
Zhenyu Liao, Romain Couillet

Figure 1 for On the Spectrum of Random Features Maps of High Dimensional Data
Figure 2 for On the Spectrum of Random Features Maps of High Dimensional Data
Figure 3 for On the Spectrum of Random Features Maps of High Dimensional Data
Figure 4 for On the Spectrum of Random Features Maps of High Dimensional Data
* 13 pages (with Supplementary Material), 10 figure, ICML 2018 
  Access Paper or Ask Questions

The Dynamics of Learning: A Random Matrix Approach


Jul 20, 2018
Zhenyu Liao, Romain Couillet

Figure 1 for The Dynamics of Learning: A Random Matrix Approach
Figure 2 for The Dynamics of Learning: A Random Matrix Approach
Figure 3 for The Dynamics of Learning: A Random Matrix Approach
Figure 4 for The Dynamics of Learning: A Random Matrix Approach
* 14 pages (with Supplementary Material), 7 figures, ICML 2018 
  Access Paper or Ask Questions

A Random Matrix Approach to Neural Networks


Jun 29, 2017
Cosme Louart, Zhenyu Liao, Romain Couillet

Figure 1 for A Random Matrix Approach to Neural Networks
Figure 2 for A Random Matrix Approach to Neural Networks
Figure 3 for A Random Matrix Approach to Neural Networks
Figure 4 for A Random Matrix Approach to Neural Networks

  Access Paper or Ask Questions

A Large Dimensional Analysis of Least Squares Support Vector Machines


Jan 11, 2017
Zhenyu Liao, Romain Couillet

Figure 1 for A Large Dimensional Analysis of Least Squares Support Vector Machines
Figure 2 for A Large Dimensional Analysis of Least Squares Support Vector Machines
Figure 3 for A Large Dimensional Analysis of Least Squares Support Vector Machines
Figure 4 for A Large Dimensional Analysis of Least Squares Support Vector Machines
* 26 pages, 10 figures, 1 table, partially presented at ICASSP 2017 
  Access Paper or Ask Questions

Random matrices meet machine learning: a large dimensional analysis of LS-SVM


Sep 08, 2016
Zhenyu Liao, Romain Couillet

Figure 1 for Random matrices meet machine learning: a large dimensional analysis of LS-SVM
Figure 2 for Random matrices meet machine learning: a large dimensional analysis of LS-SVM
Figure 3 for Random matrices meet machine learning: a large dimensional analysis of LS-SVM
Figure 4 for Random matrices meet machine learning: a large dimensional analysis of LS-SVM
* wrong article submitted 
  Access Paper or Ask Questions

Spectral Sparsification and Regret Minimization Beyond Matrix Multiplicative Updates


Jun 16, 2015
Zeyuan Allen-Zhu, Zhenyu Liao, Lorenzo Orecchia

Figure 1 for Spectral Sparsification and Regret Minimization Beyond Matrix Multiplicative Updates

  Access Paper or Ask Questions