Annual Report 2008.pdf - SAMSI

Jennifer Hoeting
Colorado State Univeristy
jah@lamar.colostate.edu
“Bayesian Hierarchical Modeling”
In this talk I will introduce the fundamentals of Bayesian hierarchical models. I will overview
Markov chain Monte Carlo methods used for estimation in Bayesian statistics. I will also discuss
the incorporation of mathematical (process) models into the hierarchy. Examples will be based
on ecological problems. If time permits, more complex hierarchical models will be discussed.
Ali Jadbabaie
University of Pennsylvania
jadbabai@seas.upenn.edu
“From Distributed Coordination and Consensus in Multi-agent Systems to Coverage in Mobile
sensor Networks”
In this talk we provide a unified view of several distributed coordination and consensus
algorithms which have appeared in various disciplines such as distributed systems, statistical
physics, biology, computer graphics, robotics, and control theory over the past decade.
These algorithms have been proposed as a mechanism for demonstrating emergence of a global
collective behavior (such as social aggregation in animals, schooling, flocking and
synchronization) using purely local interactions. Utilizing tools from spectral graph theory and
control and dynamical systems, we provide an analysis of these algorithms. Next, we extend our
results from graphs to simplicial complexes (objects of study in algebraic topology) to verify
coverage in mobile sensor networks in a purely decentralized fashion. Simplicial complexes are
combinatorial objects that generalize the proximity graphs formed from binary relations
between agents to higher order relations, and their study will allow us to infer the coverage
properties of mobile sensor networks with time-varying interconnections without localization.
The proposed approach is based on distributed computation of sparse generators of homology
groups of simplicial complexes using higher order Laplacian operators.
Bill Kaiser
University of California, Los Angeles
kaiser@ee.ucla.edu
“Sensor Network Platforms for Rapidly Deployable, Configurable, and Sustainable
Observatories”
Progress in the collaborative development of environmental science and sensor network
technology has revealed previously unanticipated, critical demands and constraints along with
new applications and results. This has led to the development of a new architecture that has now
been proven in terrestrial ecosystem and water resource monitoring. First, ecological applications
of sensor network systems require high precision sensor and instrument systems. For example,