Fast Dimension Reduction

Size: px
Start display at page:

Download "Fast Dimension Reduction"

Transcription

1 Fast Dimension Reduction Nir Ailon 1 Edo Liberty 2 1 Google Research 2 Yale University

2 Introduction Lemma (Johnson, Lindenstrauss (1984)) A random projection Ψ preserves all ( n 2) distances up to distortion ε with constant probability if: ( ) log(n) k = Ω ε 2

3 Applications This idea is extremely useful in Approximate nearest neighbors searches Linear Embedding / Dimensionality reduction Matrix rank-k approximation l p regression Compressed sensing and the list continues...

4 JL Property definition Pr[distortion] 1/n 2 embedding an n point metric. Definition A distribution D over k d matrices exhibits the JL Property if for x 2 = 1 and 0 < ε < 1/2 Pr [ Ψx 2 1 > ε] c 1e c 2kε 2 Ψ D k = c log(n)/ε 2 Pr[distortion] 1/n 2.

5 Unstructured constructions Other JL distributions for Ψ(i, j) being i.i.d random variables: Frankl and Maehara 1987 Indyk and Motwani 1998 Ψ(i, j) N(0, 1) DasGupta and Gupta 1999 Achlioptas 2003 Ψ(i, j) {0, 1, 1} Matousek 2006 Ψ(i, j) Symmetrically subgaussian distributed. However, these matrices are Slow to apply: O(kd) Heavy to store: O(kd)

6 Fast Johnson Lindenstrauss Transform The first JL distribution for efficiently applicable matrices. Lemma (Ailon, Chazelle (2006)) The distribution Ψ = PHD exhibits the JL property O(d log(d) + k 3 ) to apply. O(d + k 3 log(d)) to store.

7 Our motivation FJLT algorithm running time is O(d log(d) + k 3 ). 1. When the target dimension is very small k = O(log(d)) The running time is Ω(kd). 2. When the target dimension is large k = Ω((d log(d)) 1/3 ) The running time is dominated by k 3.

8 This work, Faster Johnson Lindenstrauss Transform Lemma (Ailon, Liberty (2008)) The distribution Ψ = BDΦ exhibits the JL property for k = O(d 1/2 δ ) and δ > 0 O(d log(k)) to apply O(d) to store.

9 Performance summary Slow Fast k in o(log d) FJLT JL This work k in ω(log d) and o(poly(d)) JL FJLT This work k in Ω(poly(d)) and o((d log(d) 1/3 ) JL This work, FJLT k in ω((d log d) 1/3 ) and O(d 1/2 δ ) JL FJLT This work

10 Measure concentration The randomness is only in D. How does Ms 2 concentrate around 1? ( x 2 = 1)

11 Sufficient conditions Lemma (Talagrand) Denote: µ the median of Ms 2, σ = M 2 2 the spectral norm of M. Pr [ Ms 2 µ > t] 4e t2 /8σ 2 By substitution we get the JL property for BD: If: Pr( BDx 2 1 > ε) c 1 e c 2kε 2 The columns of B are normalized to 1. σ = M 2 2 = O(k 1/2 )

12 Breaking down M 2 2 Using the definition of the spectral norm and Cauchy-Schwartz: M 2 2 B T 2 4 x 4 The l 2 l 4 operator norm is defined as such: B T 2 4 max z 2 =1 BT z 4 We seek a matrix B with a low l 2 l 4 operator norm, B T 2 4. which is fast to apply.

13 Our choice for B Lemma B is obtained from k rows of a Θ(k 2 ) Θ(k 2 ) Hadamard matrix. B can be applied in O(d log(k)) operations. Lemma Let B be a 4-wise independent matrix. (The columns of B are the sample space) B T 2 4 = O(d 1/4 k 1/2 ).

14 Controlling x 4 Since B T 2 4 = O(d 1/4 k 1/2 ) our goal: B T 2 4 x 4 = O(k 1/2 ) is achieved if x 4 = O(d 1/4 ) Problem: x 4 might be as high as 1! Solution: Preprocess x using a random Isometry Φ such that: Φx 4 = O(d 1/4 ) w.h.p Φ can be applied in O(d log(k)) operations Φ requires O(d) bits to store Lemma Such matrix distributions exist for k = O(d 1/2 δ ).

15 Recap and summary Any matrix distribution BDΦ exhibits the JL property if The columns of B are normalized to 1. Φ is an isometry. B T 2 4 Φx 4 O(k 1/2 ) w.h.p. Our result: For k = O(d 1/2 δ ) there exist B and Φ such that BDΦ requires O(d log(k)) operations to apply BDΦ requires O(d) bits to store

16 Future work and open questions Slow Fast O(d) k in o(log d) FJLT JL This work k in ω(log d) and o(poly(d)) JL FJLT This work k in Ω(poly(d)) and o((d log(d) 1/3 ) JL This work, FJLT k in ω((d log d) 1/3 ) and O(d 1/2 δ ) JL FJLT This work k in ω(d 1/2 δ ) and d JL, FJLT

17 Thank you

18 4-wise independence Definition B is 4-wise independent if: There are d/16 columns j for which ( B (j) (j) (j) (j) i 1, B i 2, B i 3, B i 4 ) = (b 1, b 2, b 3, b 4 ) for any choice of 1 i 1 < i 2 < i 3 < i 4 k and (b 1, b 2, b 3, b 4 ) {+1, 1} 4.

19 Bounding B T 2 4 Lemma Given a k d 4-wise independent code matrix B, B T 2 4 = O(d 1/4 k 1/2 ) (1) Proof. For y l k 2, y = 1, y T B 4 4 = de j [d][(y T B(j)) 4 ] = dk 2 k i 1,i 2,i 3,i 4 =1 E b1,b 2,b 3,b 4 [y i1 y i2 y i3 y i4 b 1 b 2 b 3 b 4 ] = dk 2 (3 y y ) 3dk (2)

20 Trimming the Hadamard transform Applying k rows from a Hadamard matrix, H, is like applying PH where P contains only k entrees which are 1 and the rest are zero. PH d z = ( ) ( ) ( ) H P 1 P d/2 H d/2 z1 2 H d/2 H d/2 z 2 = P 1 H d/2 (z 1 + z 2 ) + P 2 H d/2 (z 1 z 2 ) Which gives the relation T (d, k) = T (d/2, k 1 ) + T (d/2, k 2 ) + d. Since T (d, 1) = d, by induction T (d, k) 2d log(k + 1) = O(d log(k)) (3)

Fast Random Projections

Fast Random Projections Fast Random Projections Edo Liberty 1 September 18, 2007 1 Yale University, New Haven CT, supported by AFOSR and NGA (www.edoliberty.com) Advised by Steven Zucker. About This talk will survey a few random

More information

Accelerated Dense Random Projections

Accelerated Dense Random Projections 1 Advisor: Steven Zucker 1 Yale University, Department of Computer Science. Dimensionality reduction (1 ε) xi x j 2 Ψ(xi ) Ψ(x j ) 2 (1 + ε) xi x j 2 ( n 2) distances are ε preserved Target dimension k

More information

Fast Dimension Reduction

Fast Dimension Reduction Fast Dimension Reduction MMDS 2008 Nir Ailon Google Research NY Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes (with Edo Liberty) The Fast Johnson Lindenstrauss Transform (with Bernard

More information

Edo Liberty. Accelerated Dense Random Projections

Edo Liberty. Accelerated Dense Random Projections Edo Liberty Accelerated Dense Random Projections CONTENTS 1. List of common notations......................................... 5 2. Introduction to random projections....................................

More information

Fast Random Projections using Lean Walsh Transforms Yale University Technical report #1390

Fast Random Projections using Lean Walsh Transforms Yale University Technical report #1390 Fast Random Projections using Lean Walsh Transforms Yale University Technical report #1390 Edo Liberty Nir Ailon Amit Singer Abstract We present a k d random projection matrix that is applicable to vectors

More information

Dense Fast Random Projections and Lean Walsh Transforms

Dense Fast Random Projections and Lean Walsh Transforms Dense Fast Random Projections and Lean Walsh Transforms Edo Liberty, Nir Ailon, and Amit Singer Abstract. Random projection methods give distributions over k d matrices such that if a matrix Ψ (chosen

More information

Some Useful Background for Talk on the Fast Johnson-Lindenstrauss Transform

Some Useful Background for Talk on the Fast Johnson-Lindenstrauss Transform Some Useful Background for Talk on the Fast Johnson-Lindenstrauss Transform Nir Ailon May 22, 2007 This writeup includes very basic background material for the talk on the Fast Johnson Lindenstrauss Transform

More information

Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes

Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes Nir Ailon Edo Liberty Abstract The Fast Johnson-Lindenstrauss Transform (FJLT) was recently discovered by Ailon and Chazelle as a novel

More information

Faster Johnson-Lindenstrauss style reductions

Faster Johnson-Lindenstrauss style reductions Faster Johnson-Lindenstrauss style reductions Aditya Menon August 23, 2007 Outline 1 Introduction Dimensionality reduction The Johnson-Lindenstrauss Lemma Speeding up computation 2 The Fast Johnson-Lindenstrauss

More information

Yale University Department of Computer Science

Yale University Department of Computer Science Yale University Department of Computer Science Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes Nir Ailon 1 Edo Liberty 2 YALEU/DCS/TR-1385 July 2007 1 Institute for Advanced Study, Princeton

More information

Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes

Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes DOI 10.1007/s005-008-9110-x Fast Dimension Reduction Using Rademacher Series on Dual BCH Codes Nir Ailon Edo Liberty Received: 8 November 2007 / Revised: 9 June 2008 / Accepted: September 2008 Springer

More information

Sparser Johnson-Lindenstrauss Transforms

Sparser Johnson-Lindenstrauss Transforms Sparser Johnson-Lindenstrauss Transforms Jelani Nelson Princeton February 16, 212 joint work with Daniel Kane (Stanford) Random Projections x R d, d huge store y = Sx, where S is a k d matrix (compression)

More information

Sparse Johnson-Lindenstrauss Transforms

Sparse Johnson-Lindenstrauss Transforms Sparse Johnson-Lindenstrauss Transforms Jelani Nelson MIT May 24, 211 joint work with Daniel Kane (Harvard) Metric Johnson-Lindenstrauss lemma Metric JL (MJL) Lemma, 1984 Every set of n points in Euclidean

More information

Randomized Algorithms

Randomized Algorithms Randomized Algorithms Saniv Kumar, Google Research, NY EECS-6898, Columbia University - Fall, 010 Saniv Kumar 9/13/010 EECS6898 Large Scale Machine Learning 1 Curse of Dimensionality Gaussian Mixture Models

More information

Random Methods for Linear Algebra

Random Methods for Linear Algebra Gittens gittens@acm.caltech.edu Applied and Computational Mathematics California Institue of Technology October 2, 2009 Outline The Johnson-Lindenstrauss Transform 1 The Johnson-Lindenstrauss Transform

More information

JOHNSON-LINDENSTRAUSS TRANSFORMATION AND RANDOM PROJECTION

JOHNSON-LINDENSTRAUSS TRANSFORMATION AND RANDOM PROJECTION JOHNSON-LINDENSTRAUSS TRANSFORMATION AND RANDOM PROJECTION LONG CHEN ABSTRACT. We give a brief survey of Johnson-Lindenstrauss lemma. CONTENTS 1. Introduction 1 2. JL Transform 4 2.1. An Elementary Proof

More information

Dimensionality reduction: Johnson-Lindenstrauss lemma for structured random matrices

Dimensionality reduction: Johnson-Lindenstrauss lemma for structured random matrices Dimensionality reduction: Johnson-Lindenstrauss lemma for structured random matrices Jan Vybíral Austrian Academy of Sciences RICAM, Linz, Austria January 2011 MPI Leipzig, Germany joint work with Aicke

More information

Yale university technical report #1402.

Yale university technical report #1402. The Mailman algorithm: a note on matrix vector multiplication Yale university technical report #1402. Edo Liberty Computer Science Yale University New Haven, CT Steven W. Zucker Computer Science and Appled

More information

Optimal compression of approximate Euclidean distances

Optimal compression of approximate Euclidean distances Optimal compression of approximate Euclidean distances Noga Alon 1 Bo az Klartag 2 Abstract Let X be a set of n points of norm at most 1 in the Euclidean space R k, and suppose ε > 0. An ε-distance sketch

More information

16 Embeddings of the Euclidean metric

16 Embeddings of the Euclidean metric 16 Embeddings of the Euclidean metric In today s lecture, we will consider how well we can embed n points in the Euclidean metric (l 2 ) into other l p metrics. More formally, we ask the following question.

More information

Johnson-Lindenstrauss, Concentration and applications to Support Vector Machines and Kernels

Johnson-Lindenstrauss, Concentration and applications to Support Vector Machines and Kernels Johnson-Lindenstrauss, Concentration and applications to Support Vector Machines and Kernels Devdatt Dubhashi Department of Computer Science and Engineering, Chalmers University, dubhashi@chalmers.se Functional

More information

arxiv: v1 [cs.ds] 30 May 2010

arxiv: v1 [cs.ds] 30 May 2010 ALMOST OPTIMAL UNRESTRICTED FAST JOHNSON-LINDENSTRAUSS TRANSFORM arxiv:005.553v [cs.ds] 30 May 200 NIR AILON AND EDO LIBERTY Abstract. The problems of random projections and sparse reconstruction have

More information

Lecture 13 October 6, Covering Numbers and Maurey s Empirical Method

Lecture 13 October 6, Covering Numbers and Maurey s Empirical Method CS 395T: Sublinear Algorithms Fall 2016 Prof. Eric Price Lecture 13 October 6, 2016 Scribe: Kiyeon Jeon and Loc Hoang 1 Overview In the last lecture we covered the lower bound for p th moment (p > 2) and

More information

Randomized Algorithms

Randomized Algorithms Randomized Algorithms 南京大学 尹一通 Martingales Definition: A sequence of random variables X 0, X 1,... is a martingale if for all i > 0, E[X i X 0,...,X i1 ] = X i1 x 0, x 1,...,x i1, E[X i X 0 = x 0, X 1

More information

Sketching as a Tool for Numerical Linear Algebra

Sketching as a Tool for Numerical Linear Algebra Sketching as a Tool for Numerical Linear Algebra David P. Woodruff presented by Sepehr Assadi o(n) Big Data Reading Group University of Pennsylvania February, 2015 Sepehr Assadi (Penn) Sketching for Numerical

More information

CS on CS: Computer Science insights into Compresive Sensing (and vice versa) Piotr Indyk MIT

CS on CS: Computer Science insights into Compresive Sensing (and vice versa) Piotr Indyk MIT CS on CS: Computer Science insights into Compresive Sensing (and vice versa) Piotr Indyk MIT Sparse Approximations Goal: approximate a highdimensional vector x by x that is sparse, i.e., has few nonzero

More information

Supremum of simple stochastic processes

Supremum of simple stochastic processes Subspace embeddings Daniel Hsu COMS 4772 1 Supremum of simple stochastic processes 2 Recap: JL lemma JL lemma. For any ε (0, 1/2), point set S R d of cardinality 16 ln n S = n, and k N such that k, there

More information

MAT 585: Johnson-Lindenstrauss, Group testing, and Compressed Sensing

MAT 585: Johnson-Lindenstrauss, Group testing, and Compressed Sensing MAT 585: Johnson-Lindenstrauss, Group testing, and Compressed Sensing Afonso S. Bandeira April 9, 2015 1 The Johnson-Lindenstrauss Lemma Suppose one has n points, X = {x 1,..., x n }, in R d with d very

More information

MANY scientific computations, signal processing, data analysis and machine learning applications lead to large dimensional

MANY scientific computations, signal processing, data analysis and machine learning applications lead to large dimensional Low rank approximation and decomposition of large matrices using error correcting codes Shashanka Ubaru, Arya Mazumdar, and Yousef Saad 1 arxiv:1512.09156v3 [cs.it] 15 Jun 2017 Abstract Low rank approximation

More information

A fast randomized algorithm for approximating an SVD of a matrix

A fast randomized algorithm for approximating an SVD of a matrix A fast randomized algorithm for approximating an SVD of a matrix Joint work with Franco Woolfe, Edo Liberty, and Vladimir Rokhlin Mark Tygert Program in Applied Mathematics Yale University Place July 17,

More information

Fast Matrix Computations via Randomized Sampling. Gunnar Martinsson, The University of Colorado at Boulder

Fast Matrix Computations via Randomized Sampling. Gunnar Martinsson, The University of Colorado at Boulder Fast Matrix Computations via Randomized Sampling Gunnar Martinsson, The University of Colorado at Boulder Computational science background One of the principal developments in science and engineering over

More information

Non-Asymptotic Theory of Random Matrices Lecture 4: Dimension Reduction Date: January 16, 2007

Non-Asymptotic Theory of Random Matrices Lecture 4: Dimension Reduction Date: January 16, 2007 Non-Asymptotic Theory of Random Matrices Lecture 4: Dimension Reduction Date: January 16, 2007 Lecturer: Roman Vershynin Scribe: Matthew Herman 1 Introduction Consider the set X = {n points in R N } where

More information

Nearest Neighbor Preserving Embeddings

Nearest Neighbor Preserving Embeddings Nearest Neighbor Preserving Embeddings Piotr Indyk MIT Assaf Naor Microsoft Research Abstract In this paper we introduce the notion of nearest neighbor preserving embeddings. These are randomized embeddings

More information

Very Sparse Random Projections

Very Sparse Random Projections Very Sparse Random Projections Ping Li, Trevor Hastie and Kenneth Church [KDD 06] Presented by: Aditya Menon UCSD March 4, 2009 Presented by: Aditya Menon (UCSD) Very Sparse Random Projections March 4,

More information

The Fast Cauchy Transform and Faster Robust Linear Regression

The Fast Cauchy Transform and Faster Robust Linear Regression The Fast Cauchy Transform and Faster Robust Linear Regression Kenneth L Clarkson Petros Drineas Malik Magdon-Ismail Michael W Mahoney Xiangrui Meng David P Woodruff Abstract We provide fast algorithms

More information

Optimal Bounds for Johnson-Lindenstrauss Transformations

Optimal Bounds for Johnson-Lindenstrauss Transformations Journal of Machine Learning Research 9 08 - Submitted 5/8; Revised 9/8; Published 0/8 Optimal Bounds for Johnson-Lindenstrauss Transformations Michael Burr Shuhong Gao Department of Mathematical Sciences

More information

Acceleration of Randomized Kaczmarz Method

Acceleration of Randomized Kaczmarz Method Acceleration of Randomized Kaczmarz Method Deanna Needell [Joint work with Y. Eldar] Stanford University BIRS Banff, March 2011 Problem Background Setup Setup Let Ax = b be an overdetermined consistent

More information

Simple and Deterministic Matrix Sketches

Simple and Deterministic Matrix Sketches Simple and Deterministic Matrix Sketches Edo Liberty + ongoing work with: Mina Ghashami, Jeff Philips and David Woodruff. Edo Liberty: Simple and Deterministic Matrix Sketches 1 / 41 Data Matrices Often

More information

Randomness Efficient Fast-Johnson-Lindenstrauss Transform with Applications in Differential Privacy

Randomness Efficient Fast-Johnson-Lindenstrauss Transform with Applications in Differential Privacy Randomness Efficient Fast-Johnson-Lindenstrauss Transform with Applications in Differential Privacy Jalaj Upadhyay Cheriton School of Computer Science University of Waterloo jkupadhy@csuwaterlooca arxiv:14102470v6

More information

Lecture 18: March 15

Lecture 18: March 15 CS71 Randomness & Computation Spring 018 Instructor: Alistair Sinclair Lecture 18: March 15 Disclaimer: These notes have not been subjected to the usual scrutiny accorded to formal publications. They may

More information

Enabling very large-scale matrix computations via randomization

Enabling very large-scale matrix computations via randomization Enabling very large-scale matrix computations via randomization Gunnar Martinsson The University of Colorado at Boulder Students: Adrianna Gillman Nathan Halko Patrick Young Collaborators: Edo Liberty

More information

Randomized Algorithms in Linear Algebra and Applications in Data Analysis

Randomized Algorithms in Linear Algebra and Applications in Data Analysis Randomized Algorithms in Linear Algebra and Applications in Data Analysis Petros Drineas Rensselaer Polytechnic Institute Computer Science Department To access my web page: drineas Why linear algebra?

More information

Multivariate Statistics Random Projections and Johnson-Lindenstrauss Lemma

Multivariate Statistics Random Projections and Johnson-Lindenstrauss Lemma Multivariate Statistics Random Projections and Johnson-Lindenstrauss Lemma Suppose again we have n sample points x,..., x n R p. The data-point x i R p can be thought of as the i-th row X i of an n p-dimensional

More information

A fast randomized algorithm for overdetermined linear least-squares regression

A fast randomized algorithm for overdetermined linear least-squares regression A fast randomized algorithm for overdetermined linear least-squares regression Vladimir Rokhlin and Mark Tygert Technical Report YALEU/DCS/TR-1403 April 28, 2008 Abstract We introduce a randomized algorithm

More information

CS168: The Modern Algorithmic Toolbox Lecture #4: Dimensionality Reduction

CS168: The Modern Algorithmic Toolbox Lecture #4: Dimensionality Reduction CS168: The Modern Algorithmic Toolbox Lecture #4: Dimensionality Reduction Tim Roughgarden & Gregory Valiant April 12, 2017 1 The Curse of Dimensionality in the Nearest Neighbor Problem Lectures #1 and

More information

Similarity searching, or how to find your neighbors efficiently

Similarity searching, or how to find your neighbors efficiently Similarity searching, or how to find your neighbors efficiently Robert Krauthgamer Weizmann Institute of Science CS Research Day for Prospective Students May 1, 2009 Background Geometric spaces and techniques

More information

arxiv: v3 [cs.ds] 21 Mar 2013

arxiv: v3 [cs.ds] 21 Mar 2013 Low-distortion Subspace Embeddings in Input-sparsity Time and Applications to Robust Linear Regression Xiangrui Meng Michael W. Mahoney arxiv:1210.3135v3 [cs.ds] 21 Mar 2013 Abstract Low-distortion subspace

More information

Sketching as a Tool for Numerical Linear Algebra

Sketching as a Tool for Numerical Linear Algebra Foundations and Trends R in Theoretical Computer Science Vol. 10, No. 1-2 (2014) 1 157 c 2014 D. P. Woodruff DOI: 10.1561/0400000060 Sketching as a Tool for Numerical Linear Algebra David P. Woodruff IBM

More information

A fast randomized algorithm for the approximation of matrices preliminary report

A fast randomized algorithm for the approximation of matrices preliminary report DRAFT A fast randomized algorithm for the approximation of matrices preliminary report Yale Department of Computer Science Technical Report #1380 Franco Woolfe, Edo Liberty, Vladimir Rokhlin, and Mark

More information

Randomized methods for computing the Singular Value Decomposition (SVD) of very large matrices

Randomized methods for computing the Singular Value Decomposition (SVD) of very large matrices Randomized methods for computing the Singular Value Decomposition (SVD) of very large matrices Gunnar Martinsson The University of Colorado at Boulder Students: Adrianna Gillman Nathan Halko Sijia Hao

More information

Graph Sparsification I : Effective Resistance Sampling

Graph Sparsification I : Effective Resistance Sampling Graph Sparsification I : Effective Resistance Sampling Nikhil Srivastava Microsoft Research India Simons Institute, August 26 2014 Graphs G G=(V,E,w) undirected V = n w: E R + Sparsification Approximate

More information

Methods for sparse analysis of high-dimensional data, II

Methods for sparse analysis of high-dimensional data, II Methods for sparse analysis of high-dimensional data, II Rachel Ward May 23, 2011 High dimensional data with low-dimensional structure 300 by 300 pixel images = 90, 000 dimensions 2 / 47 High dimensional

More information

Randomized algorithms for the low-rank approximation of matrices

Randomized algorithms for the low-rank approximation of matrices Randomized algorithms for the low-rank approximation of matrices Yale Dept. of Computer Science Technical Report 1388 Edo Liberty, Franco Woolfe, Per-Gunnar Martinsson, Vladimir Rokhlin, and Mark Tygert

More information

Sketching as a Tool for Numerical Linear Algebra All Lectures. David Woodruff IBM Almaden

Sketching as a Tool for Numerical Linear Algebra All Lectures. David Woodruff IBM Almaden Sketching as a Tool for Numerical Linear Algebra All Lectures David Woodruff IBM Almaden Massive data sets Examples Internet traffic logs Financial data etc. Algorithms Want nearly linear time or less

More information

Finite Metric Spaces & Their Embeddings: Introduction and Basic Tools

Finite Metric Spaces & Their Embeddings: Introduction and Basic Tools Finite Metric Spaces & Their Embeddings: Introduction and Basic Tools Manor Mendel, CMI, Caltech 1 Finite Metric Spaces Definition of (semi) metric. (M, ρ): M a (finite) set of points. ρ a distance function

More information

Improved Bounds on the Dot Product under Random Projection and Random Sign Projection

Improved Bounds on the Dot Product under Random Projection and Random Sign Projection Improved Bounds on the Dot Product under Random Projection and Random Sign Projection Ata Kabán School of Computer Science The University of Birmingham Birmingham B15 2TT, UK http://www.cs.bham.ac.uk/

More information

MIT 9.520/6.860, Fall 2017 Statistical Learning Theory and Applications. Class 19: Data Representation by Design

MIT 9.520/6.860, Fall 2017 Statistical Learning Theory and Applications. Class 19: Data Representation by Design MIT 9.520/6.860, Fall 2017 Statistical Learning Theory and Applications Class 19: Data Representation by Design What is data representation? Let X be a data-space X M (M) F (M) X A data representation

More information

Nonlinear Estimators and Tail Bounds for Dimension Reduction in l 1 Using Cauchy Random Projections

Nonlinear Estimators and Tail Bounds for Dimension Reduction in l 1 Using Cauchy Random Projections Journal of Machine Learning Research 007 Submitted 0/06; Published Nonlinear Estimators and Tail Bounds for Dimension Reduction in l Using Cauchy Random Projections Ping Li Department of Statistical Science

More information

Lecture: Expanders, in theory and in practice (2 of 2)

Lecture: Expanders, in theory and in practice (2 of 2) Stat260/CS294: Spectral Graph Methods Lecture 9-02/19/2015 Lecture: Expanders, in theory and in practice (2 of 2) Lecturer: Michael Mahoney Scribe: Michael Mahoney Warning: these notes are still very rough.

More information

New constructions of RIP matrices with fast multiplication and fewer rows

New constructions of RIP matrices with fast multiplication and fewer rows New constructions of RIP matrices with fast multiplication and fewer rows Jelani Nelson Eric Price Mary Wootters July 8, 203 Abstract In this paper, we present novel constructions of matrices with the

More information

Fast Approximation of Matrix Coherence and Statistical Leverage

Fast Approximation of Matrix Coherence and Statistical Leverage Journal of Machine Learning Research 13 (01) 3475-3506 Submitted 7/1; Published 1/1 Fast Approximation of Matrix Coherence and Statistical Leverage Petros Drineas Malik Magdon-Ismail Department of Computer

More information

Block Heavy Hitters Alexandr Andoni, Khanh Do Ba, and Piotr Indyk

Block Heavy Hitters Alexandr Andoni, Khanh Do Ba, and Piotr Indyk Computer Science and Artificial Intelligence Laboratory Technical Report -CSAIL-TR-2008-024 May 2, 2008 Block Heavy Hitters Alexandr Andoni, Khanh Do Ba, and Piotr Indyk massachusetts institute of technology,

More information

Linear Sketches A Useful Tool in Streaming and Compressive Sensing

Linear Sketches A Useful Tool in Streaming and Compressive Sensing Linear Sketches A Useful Tool in Streaming and Compressive Sensing Qin Zhang 1-1 Linear sketch Random linear projection M : R n R k that preserves properties of any v R n with high prob. where k n. M =

More information

1 Dimension Reduction in Euclidean Space

1 Dimension Reduction in Euclidean Space CSIS0351/8601: Randomized Algorithms Lecture 6: Johnson-Lindenstrauss Lemma: Dimension Reduction Lecturer: Hubert Chan Date: 10 Oct 011 hese lecture notes are supplementary materials for the lectures.

More information

Sparsity Lower Bounds for Dimensionality Reducing Maps

Sparsity Lower Bounds for Dimensionality Reducing Maps Sparsity Lower Bounds for Dimensionality Reducing Maps Jelani Nelson Huy L. Nguy ên November 5, 01 Abstract We give near-tight lower bounds for the sparsity required in several dimensionality reducing

More information

Optimal terminal dimensionality reduction in Euclidean space

Optimal terminal dimensionality reduction in Euclidean space Optimal terminal dimensionality reduction in Euclidean space Shyam Narayanan Jelani Nelson October 22, 2018 Abstract Let ε (0, 1) and X R d be arbitrary with X having size n > 1. The Johnson- Lindenstrauss

More information

Notes taken by Costis Georgiou revised by Hamed Hatami

Notes taken by Costis Georgiou revised by Hamed Hatami CSC414 - Metric Embeddings Lecture 6: Reductions that preserve volumes and distance to affine spaces & Lower bound techniques for distortion when embedding into l Notes taken by Costis Georgiou revised

More information

Sketching as a Tool for Numerical Linear Algebra All Lectures. David Woodruff IBM Almaden

Sketching as a Tool for Numerical Linear Algebra All Lectures. David Woodruff IBM Almaden Sketching as a Tool for Numerical Linear Algebra All Lectures David Woodruff IBM Almaden Massive data sets Examples Internet traffic logs Financial data etc. Algorithms Want nearly linear time or less

More information

THE FAST CAUCHY TRANSFORM AND FASTER ROBUST LINEAR REGRESSION

THE FAST CAUCHY TRANSFORM AND FASTER ROBUST LINEAR REGRESSION SIAM J COMPUT Vol 45, No 3, pp 763 80 c 206 Society for Industrial and Applied Mathematics THE FAST CAUCHY TRANSFORM AND FASTER ROBUST LINEAR REGRESSION KENNETH L CLARKSON, PETROS DRINEAS, MALIK MAGDON-ISMAIL,

More information

Approximation norms and duality for communication complexity lower bounds

Approximation norms and duality for communication complexity lower bounds Approximation norms and duality for communication complexity lower bounds Troy Lee Columbia University Adi Shraibman Weizmann Institute From min to max The cost of a best algorithm is naturally phrased

More information

Lean Walsh Transform

Lean Walsh Transform Lean Walsh Transfor Edo Liberty 5th March 007 inforal intro We show an orthogonal atrix A of size d log 4 3 d (α = log 4 3) which is applicable in tie O(d). By applying a rando sign change atrix S to the

More information

The Johnson-Lindenstrauss Lemma for Linear and Integer Feasibility

The Johnson-Lindenstrauss Lemma for Linear and Integer Feasibility The Johnson-Lindenstrauss Lemma for Linear and Integer Feasibility Leo Liberti, Vu Khac Ky, Pierre-Louis Poirion CNRS LIX Ecole Polytechnique, France Rutgers University, 151118 The gist I want to solve

More information

Proximity problems in high dimensions

Proximity problems in high dimensions Proximity problems in high dimensions Ioannis Psarros National & Kapodistrian University of Athens March 31, 2017 Ioannis Psarros Proximity problems in high dimensions March 31, 2017 1 / 43 Problem definition

More information

An Approximation Algorithm for Approximation Rank

An Approximation Algorithm for Approximation Rank An Approximation Algorithm for Approximation Rank Troy Lee Columbia University Adi Shraibman Weizmann Institute Conventions Identify a communication function f : X Y { 1, +1} with the associated X-by-Y

More information

Similarity Search in High Dimensions II. Piotr Indyk MIT

Similarity Search in High Dimensions II. Piotr Indyk MIT Similarity Search in High Dimensions II Piotr Indyk MIT Approximate Near(est) Neighbor c-approximate Nearest Neighbor: build data structure which, for any query q returns p P, p-q cr, where r is the distance

More information

CS 229r: Algorithms for Big Data Fall Lecture 19 Nov 5

CS 229r: Algorithms for Big Data Fall Lecture 19 Nov 5 CS 229r: Algorithms for Big Data Fall 215 Prof. Jelani Nelson Lecture 19 Nov 5 Scribe: Abdul Wasay 1 Overview In the last lecture, we started discussing the problem of compressed sensing where we are given

More information

Low Rank Matrix Approximation

Low Rank Matrix Approximation Low Rank Matrix Approximation John T. Svadlenka Ph.D. Program in Computer Science The Graduate Center of the City University of New York New York, NY 10036 USA jsvadlenka@gradcenter.cuny.edu Abstract Low

More information

A Unified Theorem on SDP Rank Reduction. yyye

A Unified Theorem on SDP Rank Reduction.   yyye SDP Rank Reduction Yinyu Ye, EURO XXII 1 A Unified Theorem on SDP Rank Reduction Yinyu Ye Department of Management Science and Engineering and Institute of Computational and Mathematical Engineering Stanford

More information

An Algorithmist s Toolkit Nov. 10, Lecture 17

An Algorithmist s Toolkit Nov. 10, Lecture 17 8.409 An Algorithmist s Toolkit Nov. 0, 009 Lecturer: Jonathan Kelner Lecture 7 Johnson-Lindenstrauss Theorem. Recap We first recap a theorem (isoperimetric inequality) and a lemma (concentration) from

More information

17 Random Projections and Orthogonal Matching Pursuit

17 Random Projections and Orthogonal Matching Pursuit 17 Random Projections and Orthogonal Matching Pursuit Again we will consider high-dimensional data P. Now we will consider the uses and effects of randomness. We will use it to simplify P (put it in a

More information

Norms of Random Matrices & Low-Rank via Sampling

Norms of Random Matrices & Low-Rank via Sampling CS369M: Algorithms for Modern Massive Data Set Analysis Lecture 4-10/05/2009 Norms of Random Matrices & Low-Rank via Sampling Lecturer: Michael Mahoney Scribes: Jacob Bien and Noah Youngs *Unedited Notes

More information

18.409: Topics in TCS: Embeddings of Finite Metric Spaces. Lecture 6

18.409: Topics in TCS: Embeddings of Finite Metric Spaces. Lecture 6 Massachusetts Institute of Technology Lecturer: Michel X. Goemans 8.409: Topics in TCS: Embeddings of Finite Metric Spaces September 7, 006 Scribe: Benjamin Rossman Lecture 6 Today we loo at dimension

More information

OSNAP: Faster numerical linear algebra algorithms via sparser subspace embeddings

OSNAP: Faster numerical linear algebra algorithms via sparser subspace embeddings OSNAP: Faster numerical linear algebra algorithms via sparser subspace embeddings Jelani Nelson Huy L. Nguy ên November 5, 2012 Abstract An oblivious subspace embedding (OSE) given some parameters ε, d

More information

Randomized Numerical Linear Algebra: Review and Progresses

Randomized Numerical Linear Algebra: Review and Progresses ized ized SVD ized : Review and Progresses Zhihua Department of Computer Science and Engineering Shanghai Jiao Tong University The 12th China Workshop on Machine Learning and Applications Xi an, November

More information

Lecture 16: Compressed Sensing

Lecture 16: Compressed Sensing Lecture 16: Compressed Sensing Introduction to Learning and Analysis of Big Data Kontorovich and Sabato (BGU) Lecture 16 1 / 12 Review of Johnson-Lindenstrauss Unsupervised learning technique key insight:

More information

1 Singular Value Decomposition and Principal Component

1 Singular Value Decomposition and Principal Component Singular Value Decomposition and Principal Component Analysis In these lectures we discuss the SVD and the PCA, two of the most widely used tools in machine learning. Principal Component Analysis (PCA)

More information

Optimal Data-Dependent Hashing for Approximate Near Neighbors

Optimal Data-Dependent Hashing for Approximate Near Neighbors Optimal Data-Dependent Hashing for Approximate Near Neighbors Alexandr Andoni 1 Ilya Razenshteyn 2 1 Simons Institute 2 MIT, CSAIL April 20, 2015 1 / 30 Nearest Neighbor Search (NNS) Let P be an n-point

More information

Lecture 6 Proof for JL Lemma and Linear Dimensionality Reduction

Lecture 6 Proof for JL Lemma and Linear Dimensionality Reduction COMS 4995: Unsupervised Learning (Summer 18) June 7, 018 Lecture 6 Proof for JL Lemma and Linear imensionality Reduction Instructor: Nakul Verma Scribes: Ziyuan Zhong, Kirsten Blancato This lecture gives

More information

Dimensionality reduction of SDPs through sketching

Dimensionality reduction of SDPs through sketching Technische Universität München Workshop on "Probabilistic techniques and Quantum Information Theory", Institut Henri Poincaré Joint work with Andreas Bluhm arxiv:1707.09863 Semidefinite Programs (SDPs)

More information

Methods for sparse analysis of high-dimensional data, II

Methods for sparse analysis of high-dimensional data, II Methods for sparse analysis of high-dimensional data, II Rachel Ward May 26, 2011 High dimensional data with low-dimensional structure 300 by 300 pixel images = 90, 000 dimensions 2 / 55 High dimensional

More information

Rank minimization via the γ 2 norm

Rank minimization via the γ 2 norm Rank minimization via the γ 2 norm Troy Lee Columbia University Adi Shraibman Weizmann Institute Rank Minimization Problem Consider the following problem min X rank(x) A i, X b i for i = 1,..., k Arises

More information

sublinear time low-rank approximation of positive semidefinite matrices Cameron Musco (MIT) and David P. Woodru (CMU)

sublinear time low-rank approximation of positive semidefinite matrices Cameron Musco (MIT) and David P. Woodru (CMU) sublinear time low-rank approximation of positive semidefinite matrices Cameron Musco (MIT) and David P. Woodru (CMU) 0 overview Our Contributions: 1 overview Our Contributions: A near optimal low-rank

More information

sample lectures: Compressed Sensing, Random Sampling I & II

sample lectures: Compressed Sensing, Random Sampling I & II P. Jung, CommIT, TU-Berlin]: sample lectures Compressed Sensing / Random Sampling I & II 1/68 sample lectures: Compressed Sensing, Random Sampling I & II Peter Jung Communications and Information Theory

More information

Advances in Manifold Learning Presented by: Naku Nak l Verm r a June 10, 2008

Advances in Manifold Learning Presented by: Naku Nak l Verm r a June 10, 2008 Advances in Manifold Learning Presented by: Nakul Verma June 10, 008 Outline Motivation Manifolds Manifold Learning Random projection of manifolds for dimension reduction Introduction to random projections

More information

Applied Analysis (APPM 5440): Final exam 1:30pm 4:00pm, Dec. 14, Closed books.

Applied Analysis (APPM 5440): Final exam 1:30pm 4:00pm, Dec. 14, Closed books. Applied Analysis APPM 44: Final exam 1:3pm 4:pm, Dec. 14, 29. Closed books. Problem 1: 2p Set I = [, 1]. Prove that there is a continuous function u on I such that 1 ux 1 x sin ut 2 dt = cosx, x I. Define

More information

High Dimensional Geometry, Curse of Dimensionality, Dimension Reduction

High Dimensional Geometry, Curse of Dimensionality, Dimension Reduction Chapter 11 High Dimensional Geometry, Curse of Dimensionality, Dimension Reduction High-dimensional vectors are ubiquitous in applications (gene expression data, set of movies watched by Netflix customer,

More information

A Unifying Framework for l 0 -Sampling Algorithms

A Unifying Framework for l 0 -Sampling Algorithms Distributed and Parallel Databases manuscript No. (will be inserted by the editor) A Unifying Framework for l 0 -Sampling Algorithms Graham Cormode Donatella Firmani Received: date / Accepted: date Abstract

More information

Compressed Sensing: Lecture I. Ronald DeVore

Compressed Sensing: Lecture I. Ronald DeVore Compressed Sensing: Lecture I Ronald DeVore Motivation Compressed Sensing is a new paradigm for signal/image/function acquisition Motivation Compressed Sensing is a new paradigm for signal/image/function

More information

EECS 275 Matrix Computation

EECS 275 Matrix Computation EECS 275 Matrix Computation Ming-Hsuan Yang Electrical Engineering and Computer Science University of California at Merced Merced, CA 95344 http://faculty.ucmerced.edu/mhyang Lecture 22 1 / 21 Overview

More information

A Simple Proof of the Restricted Isometry Property for Random Matrices

A Simple Proof of the Restricted Isometry Property for Random Matrices DOI 10.1007/s00365-007-9003-x A Simple Proof of the Restricted Isometry Property for Random Matrices Richard Baraniuk Mark Davenport Ronald DeVore Michael Wakin Received: 17 May 006 / Revised: 18 January

More information