Low-rank approximations

Give matrix and a positive integer , we wish to find an matrix of rank at most ，so as to minimize the Frobenius norm of the matrix difference ，defined to be

Thus, the Frobenius norm of measures the discrepancy between and ; our goal is to find a matrix that minimizes this discrepancy, while constraining to have rank at most . If is the rank of , clearly and the Frobenius norm of the discrepancy is zero in this case. When is far smaller than , we refer to as a low-rank approximation.

SVD

  1. Given , constuct its SVD int the form of .
  2. Derive from the matrix formed by replacing by zeros the smallest sigular values on the diagnal of .
  3. Compute and output as the rank- approximation to .

http://nlp.stanford.edu/IR-book/html/htmledition/low-rank-approximations-1.html

Low-rank approximations的更多相关文章

  1. Generalized Low Rank Approximation of Matrices

    Generalized Low Rank Approximations of Matrices JIEPING YE*jieping@cs.umn.edu Department of Computer ...

  2. 吴恩达机器学习笔记59-向量化：低秩矩阵分解与均值归一化（Vectorization: Low Rank Matrix Factorization & Mean Normalization）

    一.向量化:低秩矩阵分解 之前我们介绍了协同过滤算法,本节介绍该算法的向量化实现,以及说说有关该算法可以做的其他事情. 举例:1.当给出一件产品时,你能否找到与之相关的其它产品.2.一位用户最近看上一 ...

  3. Sparse Principal Component Analysis via Regularized Low Rank Matrix Approximation(Adjusted Variance)

    目录 前言 文章概述 固定\(\widetilde{\mathrm{v}}\) 固定\(\widetilde{\mathrm{u}}\) Adjusted Variance 前言 这篇文章用的也是交替 ...

  4. 推荐系统(recommender systems):预测电影评分--构造推荐系统的一种方法：低秩矩阵分解(low rank matrix factorization)

    如上图中的predicted ratings矩阵可以分解成X与ΘT的乘积,这个叫做低秩矩阵分解. 我们先学习出product的特征参数向量,在实际应用中这些学习出来的参数向量可能比较难以理解,也很难可 ...

  5. 矩阵分解（rank decomposition）文章代码汇总

    矩阵分解(rank decomposition)文章代码汇总 矩阵分解(rank decomposition) 本文收集了现有矩阵分解的几乎所有算法和应用,原文链接:https://sites.goo ...

  6. <<Numerical Analysis>>笔记

    2ed,  by Timothy Sauer DEFINITION 1.3A solution is correct within p decimal places if the error is l ...

  7. [综述]Deep Compression/Acceleration深度压缩/加速/量化

    Survey Recent Advances in Efficient Computation of Deep Convolutional Neural Networks, [arxiv '18] A ...

  8. <Numerical Analysis>(by Timothy Sauer) Notes

    2ed,  by Timothy Sauer DEFINITION 1.3A solution is correct within p decimal places if the error is l ...

  9. 网络压缩论文集(network compression)

    Convolutional Neural Networks ImageNet Models Architecture Design Activation Functions Visualization ...

  10. CVPR 2015 papers

    CVPR2015 Papers震撼来袭! CVPR 2015的文章可以下载了,如果链接无法下载,可以在Google上通过搜索paper名字下载(友情提示:可以使用filetype:pdf命令). Go ...

随机推荐

  1. Java知IO

    ---恢复内容开始--- Java将IO(文件.网络.终端)封装成非常多的类,看似繁杂,其实每个类的具有独特的功能. 按照存取的对象是二进制还是文本,java使用字节流和字符流实现IO. 流是java ...

  2. [LeetCode] String Compression 字符串压缩

    Given an array of characters, compress it in-place. The length after compression must always be smal ...

  3. [LeetCode] Split Linked List in Parts 拆分链表成部分

    Given a (singly) linked list with head node root, write a function to split the linked list into k c ...

  4. 学习flexible.js

    flexible是手淘团队总结的一套移动端适配方案,这篇文章是自己对flexible学习历程的记录,也算是一个备忘. https://github.com/amfe/article/issues/17 ...

  5. Virtual Box下虚拟机复制后ip地址重复

    通过桥接模式上网的虚拟机在复制之后,出现三台机器的ip地址都是一样的,还都可以上网, 主要是因为在复制的时候,把网卡信息啥的都一起复制了, 为了设置为不同的ip,需要修改复制后的机器的mac地址. 首 ...

  6. [HNOI 2013]游走

    Description 题库链接 一个无向连通图,顶点从 \(1\) 编号到 \(N\) ,边从 \(1\) 编号到 \(M\) . 小Z在该图上进行随机游走,初始时小Z在 \(1\) 号顶点,每一步 ...

  7. [TJOI 2017]可乐

    Description 加里敦星球的人们特别喜欢喝可乐.因而,他们的敌对星球研发出了一个可乐机器人,并且放在了加里敦星球的1号城市上.这个可乐机器人有三种行为: 停在原地,去下一个相邻的城市,自爆.它 ...

  8. Prison 监狱

    [题目描述]Caima 王国中有一个奇怪的监狱,这个监狱一共有 P 个牢房,这些牢房一字排开,第 i 个仅挨着第 i+1 个(最后一个除外).现在正好牢房是满的.上级下发了一个释放名单,要求每天释放名 ...

  9. 【BZOJ1060】【ZJOI2007】时态同步

    Description 小Q在电子工艺实习课上学习焊接电路板.一块电路板由若干个元件组成,我们不妨称之为节点,并将其用数字1,2,3-.进行标号.电路板的各个节点由若干不相交的导线相连接,且对于电路板 ...

  10. 2015 多校联赛 ——HDU5375（dp）

    Sample Input 2 00?0 1 2 4 8 ???? 1 2 4 8   Sample Output Case #1: 12 Case #2: 15 ?部分可以是0  or  1,将二进制 ...