Department of Computer Science | Institute of Theoretical Computer Science | CADMO

Prof. Emo Welzl and Prof. Bernd Gärtner

Mittagsseminar Talk Information |

**Date and Time**: Thursday, October 28, 2021, 12:15 pm

**Duration**: 30 minutes

**Location**: OAT S15/S16/S17

**Speaker**: Julian Portmann

We study the problem of finding a treasure in an unweighted undirected graph by a mobile agent with advice given in the form of pebbles. This variant of the treasure hunting problem was introduced by Gorain et al. (SIROCCO '21). The vertices are anonymous and the edges incident to a vertex v are labeled from 1 to deg(v). Without prior knowledge of the graph and starting from an arbitrary source vertex s, the agent must find a treasure located on an arbitrary unknown target vertex t in unknown distance D from s.

To speed up the exploration, an oracle that has complete knowledge of the graph, s, and t, is allowed to give advice to the agent in the form of placing pebbles on the vertices of the graph, at most one per vertex. The goal is to design such an oracle and a corresponding treasure hunting algorithm that minimizes the time needed, measured as the number of edge traversals, for the agent to find the treasure. Gorain et al. showed that Ω(D log ∆) steps are necessary and O(D log ∆ + log³ ∆) steps suffice, where ∆ is the maximum degree in the graph. In this work, we show that their lower bound is tight by providing an algorithm that yields a matching upper bound: treasure hunting can be performed in O(D log ∆) steps using O(D log ∆) pebbles as advice.

Joint work with Sebastian Brandt, Davin Choo, Christoph Grunau, and Václav Rozhoň.

Upcoming talks | All previous talks | Talks by speaker | Upcoming talks in iCal format (beta version!)

Previous talks by year: 2023 2022 2021 2020 2019 2018 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996

Information for students and suggested topics for student talks

Automatic MiSe System Software Version 1.4803M | admin login