low-rank-approximation

je-suis-tm/machine-learning

Python machine learning applications in image processing, recommender system, matrix completion, netflix problem and algorithm implementations including Co-clustering, Funk SVD, SVD++, Non-negative Matrix Factorization, Koren Neighborhood Model, Koren Integrated Model, Dawid-Skene, Platt-Burges, Expectation Maximization, Factor Analysis, ISTA, FISTA, ADMM, Gaussian Mixture Model, OPTICS, DBSCAN, Random Forest, Decision Tree, Support Vector Machine, Independent Component Analysis, Latent Semantic Indexing, Principal Component Analysis, Singular Value Decomposition, K Nearest Neighbors, K Means, Naïve Bayes Mixture Model, Gaussian Discriminant Analysis, Newton Method, Coordinate Descent, Gradient Descent, Elastic Net Regression, Ridge Regression, Lasso Regression, Least Squares, Logistic Regression, Linear Regression

lixilinx/psgd_torch

Pytorch implementation of preconditioned stochastic gradient descent (Kron and affine preconditioner, low-rank approximation preconditioner and more)

csjunxu/MCWNNM-ICCV2017

Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoising, ICCV 2017.

brian6091/Dreambooth

Fine-tuning of diffusion models

rockerBOO/lora-inspector

LoRA (Low-Rank Adaptation) inspector for Stable Diffusion

vortex-exoplanet/VIP

VIP is a python package/library for angular, reference star and spectral differential imaging for exoplanet/disk detection through high-contrast imaging.

ofsoundof/group_sparsity

Group Sparsity: The Hinge Between Filter Pruning and Decomposition for Network Compression. CVPR2020.

AndreiChertkov/teneva

A framework based on the tensor train decomposition for working with multivariate functions and multidimensional arrays

lixilinx/psgd_tf

Tensorflow implementation of preconditioned stochastic gradient descent

pashtari/factorizer

Pytorch implementation of Factorizer.

ecrc/hicma

HiCMA: Hierarchical Computations on Manycore Architectures

garyfanhku/Galore-pytorch

GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection

twinkle0331/Xcompression

[ICLR 2022] Code for paper "Exploring Extreme Parameter Compression for Pre-trained Language Models"(https://arxiv.org/abs/2205.10036)

ofsoundof/learning_filter_basis

Pytorch implemenation of "Learning Filter Basis for Convolutional Neural Network Compression" ICCV2019

DavisLaboratory/msImpute

Methods for label-free mass spectrometry proteomics

AndreiChertkov/fpcross

Solver in the low-rank tensor train format with cross approximation approach for the multidimensional Fokker-Planck equation

pashtari/lrf

PyTorch implementation of low-rank factorization (LRF) methods for data compression

degleris1/CMF.jl

Convolutive Matrix Factorization in Julia

UniPD-DII-ETCOMP/DenseMatrixMarket

Dense Matrix Market

loyalliu/MS-HTC

Multi-slice MR Reconstruction with Low-Rank Tensor Completion

roland1993/d_RPCA

Deformable Groupwise Image Registration using Low-Rank and Sparse Decomposition

kingofspace0wzz/multilayer_nmf

My experiment of multilayer NMF, a deep neural network in which the first several layers take Semi-NMF as its pseudo-activation-function that finds the latent sturcture embedding in the original data unsupervisely.

mpimd-csc/Structure-preserving_STTM

This repository contains MATLAB files for the implementation of work proposed in the paper Efficient Structure-preserving Support Tensor Train Machine.

musco-ai/musco-tf

MUSCO: Multi-Stage COmpression of neural networks

hi-paris/Lowrankdensity

Lowrankdensity

ecrc/stars-h

Software for Testing Accuracy, Reliability and Scalability of Hierarchical computations.

loyalliu/MS-HTC2

Calibrationless Multi-Slice Cartesian MRI via Orthogonally Alternating Phase Encoding Direction and Joint Low-Rank Tensor Completion

ztanml/arLMM

Approximate Ridge Linear Mixed Models (arLMM)

anishacharya/Online-Embedding-Compression-AAAI-2019

Deep learning models have become state of the art for natural language processing (NLP) tasks, however deploying these models in production system poses significant memory constraints. Existing compression methods are either lossy or introduce significant latency. We propose a compression method that leverages low rank matrix factorization during training, to compress the word embedding layer which represents the size bottleneck for most NLP models. Our models are trained, compressed and then further re-trained on the downstream task to recover accuracy while maintaining the reduced size. Empirically, we show that the proposed method can achieve 90% compression with minimal impact in accuracy for sentence classification tasks, and outperforms alternative methods like fixed-point quantization or offline word embedding compression. We also analyze the inference time and storage space for our method through FLOP calculations, showing that we can compress DNN models by a configurable ratio and regain accuracy loss without introducing additional latency compared to fixed point quantization. Finally, we introduce a novel learning rate schedule, the Cyclically Annealed Learning Rate (CALR), which we empirically demonstrate to outperform other popular adaptive learning rate algorithms on a sentence classification benchmark.

bangtianliu/MatRox_RU

IST-DASLab/EFCP

The repository contains code to reproduce the experiments from our paper Error Feedback Can Accurately Compress Preconditioners available below:

mzalaya/collectivemc

Implementation of Collective Matrix Completion by Mokhtar Z. Alaya and Olga Klopp https://arxiv.org/abs/1807.09010

FMatti/ACA-SPSD

Adaptive cross approximation (ACA) algorithms for symmetric positive semi-definite (SPSD) matrices.

gdikov/stochastic-segmentation-networks

Tutorial reimplementation of Monteiro et al. (2020) on a toy problem.

kaylode/rec-sys

Introducing traditional algorithms in Recommendation System.