Strong Coresets for Subspace Approximation and k-Median in Nearly Linear Time

12/27/2019
by   Zhili Feng, et al.
0

Recently the first (1+ϵ)-approximate strong coresets for k-median and subspace approximation of size poly(k/ϵ) were obtained by Sohler and Woodruff 2018. Importantly, given n points in R^d, the size of these coresets was the first that was independent of both n and d. Unfortunately their construction had a running time which was exponential in poly(k/ϵ). Here we give the first polynomial time, and in fact nearly linear time, algorithms for constructing such coresets. Our first algorithm runs in nnz(A)/ϵ + (n+d) poly(k/ϵ) time and our second runs in nd log^2(nd) + (n+d) poly(k/ϵ) time.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/09/2018

Strong Coresets for k-Median and Subspace Approximation: Goodbye Dimension

We obtain the first strong coresets for the k-median and subspace approx...
research
06/07/2019

Robust subgaussian estimation of a mean vector in nearly linear time

We construct an algorithm, running in nearly-linear time, which is robus...
research
10/15/2022

A Nearly Optimal Size Coreset Algorithm with Nearly Linear Time

A coreset is a point set containing information about geometric properti...
research
09/12/2019

Near-Linear Time Algorithms for Streett Objectives in Graphs and MDPs

The fundamental model-checking problem, given as input a model and a spe...
research
08/29/2019

Nearly Tight Bounds for Robust Proper Learning of Halfspaces with a Margin

We study the problem of properly learning large margin halfspaces in th...
research
12/29/2022

Near-Tight Algorithms for the Chamberlin-Courant and Thiele Voting Rules

We present an almost optimal algorithm for the classic Chamberlin-Couran...

Please sign up or login with your details

Forgot password? Click here to reset