In recent years, gossip-based algorithms have gained prominence as a methodology for designing robust and scalable communication schemes in large distributed systems. The premise underlying distributed gossip is very simple: in each time step, each node v in the system selects some other node w as a communication partner — generally by a simple randomized rule — and exchanges information with w; over a period of time, information spreads through the system in an “epidemic fashion”. A fundamental issue which is not well understood is the following: how does the underlying low-level gossip mechanism — the means by which communication partners are chosen — affect one’s ability to design efficient high-level gossip-based protocols? We establish one of the first concrete results addressing this question, by showing a fundamental limitation on the power of the commonly used uniform gossip mechanism for solving nearest-resource location problems. In contrast, very efficient protocols for this problem can be designed using a non-uniform spatial gossip mechanism, as established in earlier work with Alan Demers. We go on to consider the design of protocols for more complex problems, providing an efficient distributed gossipbased protocol for a set of nodes in Euclidean space to construct an approximate minimum spanning tree. Here too, we establish a contrasting limitation on the power of uniform gossip for solving this problem. Finally, we investigate gossip-based packet routing as a primitive that underpins the communication patterns in many protocols, and as a way to understand the capabilities of different gossip mechanisms at a general level.

Abstract We develop an external memory algorithm for computing minimum spanning trees. The algorithm is considerably simpler than previously known external memory algorithms for this problem and needs a factor of at least four less I/Os for realistic inputs. Our implementation indicates that this algorithm processes graphs only limited by the disk capacity of most current machines in time no more than a factor 2–5 of a good internal algorithm with sufficient memory space.

We present a hierarchical partitioning of images using a pairwise similarity function on a graph-based representation of an image. This function measures the difference along the boundary of two components relative to a measure of differences of the components' internal differences. This definition tries to encapsulate the intuitive notion of contrast. Two components are merged if there is a low-cost connection between them. Each component's internal difference is represented by the maximum edge weight of its minimum spanning tree. External differences are the smallest weight of edges connecting components. We use this idea for building a minimum spanning tree to find region borders quickly and effortlessly in a bottom-up way, based on local differences in a specific feature.

The region's internal properties (color, texture, ...) help to identify them and their external relations (adjacency, inclusion, ...) are used to build groups of regions having a particular consistent meaning in a more abstract context. Low-level cue image segmentation in a bottom-up way, cannot and should not produce a complete final "good" segmentation. We present a hierarchical partitioning of images using a pairwise similarity function on a graph-based representation of an image.

We present a hierarchical partitioning of images using a pairwise similarity function on a graph-based representation of an image.

We present Filter-Kruskal – a simple modification of Kruskal’s algorithm that avoids sorting edges that are “obviously ” not in the MST. For arbitrary graphs with random edge weights Filter-Kruskal runs in time O ( m + n lognlog m n, i.e. in linear time for not too sparse graphs. Experiments indicate that the algorithm has very good practical performance over the entire range of edge densities. An equally simple parallelization seems to be the currently best practical algorithm on multicore machines. 1

Dissertation présentée en vue de l’obtention du titre de Docteur en

The „wisdom of the crowds ‟ effect describes the finding that combining responses across a number of individuals in a group leads to aggregate performance that is as good as or better than the performance of the best individuals in the group. Here, we look at the wisdom of the crowds effect in the Minimum Spanning Tree Problem (MSTP). The MSTP is an optimization problem where observers must connect a set of nodes into a network with the shortest path length possible. A method is developed that creates aggregate solutions based only on the nodes connected in individuals ‟ solutions, without access to spatial information about the nodes. Across the three problems analyzed, the solutions produced by the aggregation method perform better than even the best individual, leading to a strong wisdom of the crowds effect. We show this effect can be observed even with sample sizes as small as 6 individuals.