Publications de Leonidas Georgopoulos

Distributed machine learning in networks by consensus

We propose an algorithm to learn from distributed data on a network of arbitrarily connected machines without exchange of the data-points. Parts of the dataset are processed locally at each machine, and then the consensus communication algorithm is employe ...

Elsevier2014

Training Distributed Neural Networks by Consensus

Martin Hasler, Leonidas Georgopoulos

We address the problem of learning a classifier from distributed data over a number of arbitrarily connected machines without exchange of the datapoints. Our purpose is to train a neural network at each machine as if the entire dataset was locally availabl ...

2011

Definitive Consensus for Distributed Data Inference

Leonidas Georgopoulos

Inference from data is of key importance in many applications of informatics. The current trend in performing such a task of inference from data is to utilise machine learning algorithms. Moreover, in many applications that it is either required or is pref ...

EPFL2011

Early consensus in complex networks under variable graph topology

Martin Hasler, Leonidas Georgopoulos

Analysis of the linear algorithm for consensus on complex networks shows the existence of two distinct phases, the asymptotic and the transient. The network reaches the asymptotic when the components corresponding to the spectral gap eigenvalue of the weig ...

IEEE2009

Nonlinear Average Consensus

Martin Hasler, Leonidas Georgopoulos

We present an algorithm for nonlinear consensus in complex networks. Our motivation draws from analysis on the algorithm based on a weighted linear update protocol. Comparison of the asymptotic with early convergence rate encourages an alternative algorith ...

IEICE2009

Forest fire localization using distributed algorithms in wireless sensor networks

Martin Hasler, Alireza Khadivi, Leonidas Georgopoulos

The fire localization using a distributed consensus finding algorithm in a wireless sensor network is described. The fire is circumscribed by a circle. The information is available at all sensor nodes that are alive, which makes it robust against failures ...

Electromagnetics Acad, 777 Concord Avenue, Ste 207, Cambridge, Ma 02138 Usa2009

Forest fire localization using a distributed consensus algorithm in wireless sensor networks

Martin Hasler, Alireza Khadivi, Leonidas Georgopoulos

2009