Max-Planck-Institut für Informatik
max planck institut
mpii logo Minerva of the Max Planck Society

MPI-INF or MPI-SWS or Local Campus Event Calendar

<< Previous Entry Next Entry >> New Event Entry Edit this Entry Login to DB (to update, delete)
What and Who
Title:Algorithmic Results for Clustering and Refined Physarum Analysis
Speaker:Pavel Kolev
coming from:Max-Planck-Institut für Informatik - D1
Speakers Bio:
Event Type:Promotionskolloquium
Visibility:D1, D3, D4, RG1, MMCI, D2, D5, SWS
We use this to send out email in the morning.
Level:Public Audience
Date, Time and Location
Date:Tuesday, 27 November 2018
Duration:60 Minutes
Building:E1 5
In the first part of this thesis, we study the Binary $\ell_0$-Rank-$k$ problem which given a binary matrix $A$ and a positive integer $k$, seeks to find a rank-$k$ binary matrix $B$ minimizing the number of non-zero entries of $A-B$. A central open question is whether this problem admits a polynomial time approximation scheme.

We give an affirmative answer to this question by designing the first randomized almost-linear time approximation scheme for constant $k$ over the reals, $\mathbb{F}_2$, and the Boolean semiring. In addition, we give novel algorithms for important variants of $\ell_0$-low rank approximation.

The second part of this dissertation, studies a popular and successful heuristic, known as Approximate Spectral Clustering (ASC), for partitioning the nodes of a graph $G$ into clusters with small conductance. We give a comprehensive analysis, showing that ASC runs efficiently and yields a good approximation of an optimal $k$-way node partition of $G$.

In the final part of this thesis, we present two results on slime mold computations: i) the continuous undirected Physarum dynamics converges for undirected linear programs with a non-negative cost vector; and ii) for the discrete directed Physarum dynamics, we give a refined analysis that yields strengthened and close to optimal convergence rate bounds, and shows that the model can be initialized with any strongly dominating point.

Name(s):Pavel Kolev
Video Broadcast
Video Broadcast:NoTo Location:
Tags, Category, Keywords and additional notes
Attachments, File(s):

Pavel Kolev, 11/26/2018 10:53 AM
Last modified:
Uwe Brahm/MPII/DE, 11/27/2018 07:01 AM
  • Pavel Kolev, 11/26/2018 10:53 AM -- Created document.