22,389 Hits in 4.0 sec

Peer-to-peer Federated Learning on Graphs [article]

Anusha Lalitha, Osman Cihan Kilinc, Tara Javidi, Farinaz Koushanfar
2019 arXiv   pre-print
Experiments on training linear regression model and on training a DNN show that the proposed learning rule algorithm provides a significant improvement in the accuracy compared to the case where nodes  ...  We propose a distributed learning algorithm in which nodes update their belief by aggregate information from their one-hop neighbors to learn a model that best fits the observations over the entire network  ...  Algorithm 1 Peer-to-peer Federated Learning Algorithm 1: Inputs: ρ (0) i ∈ P(Θ) with ρ (0) i > 0 for all i ∈ [N ] 2: Outputs:θ (n) i for all i ∈ [N ] 3: for instance k = 1 to n do 4: for node i = 1 to  ... 
arXiv:1901.11173v1 fatcat:bwbtxuapc5euplrhsgtim4texq

Scatterbrained: A flexible and expandable pattern for decentralized machine learning [article]

Miller Wilt, Jordan K. Matelsky, Andrew S. Gearhart
2021 arXiv   pre-print
One weakness of this approach is that most federated learning tools rely upon a central server to perform workload delegation and to produce a single shared model.  ...  Because data remains local to each compute node, federated learning is well-suited for use-cases in fields where data is carefully controlled, such as medicine, or in domains with bandwidth constraints  ...  The compute graph topology describes how nodes in a federated learning cluster connect to one another.  ... 
arXiv:2112.07718v1 fatcat:emgflmbiv5a2fmnbaglocofrna

Deep Federated Learning for Autonomous Driving [article]

Anh Nguyen, Tuong Do, Minh Tran, Binh X. Nguyen, Chien Duong, Tu Phan, Erman Tjiputra, Quang D. Tran
2022 arXiv   pre-print
We propose a peer-to-peer Deep Federated Learning (DFL) approach to train deep architectures in a fully decentralized manner and remove the need for central orchestration.  ...  In this paper, we present a new approach to learn autonomous driving policy while respecting privacy concerns.  ...  An overview of our peer-to-peer Deep Federated Learning method. (a) A simplified version of an overlay graph. (b) The training methodology in the overlay graph.  ... 
arXiv:2110.05754v2 fatcat:yr2gz3sd6jg2hmdavidigim4uy

p2pGNN: A Decentralized Graph Neural Network for Node Classification in Peer-to-Peer Networks [article]

Emmanouil Krasanakis, Symeon Papadopoulos, Ioannis Kompatsiaris
2021 arXiv   pre-print
We experiment on three real-world graphs with node features and labels and simulate peer-to-peer networks with uniformly random communication frequencies; given a portion of known labels, our decentralized  ...  For these, we deploy pre-trained and gossip-trained base classifiers and implement peer-to-peer graph diffusion under communication uncertainty.  ...  By definition, distributed and federated learning train one central model that is fed back to workers to make inferences.  ... 
arXiv:2111.14837v1 fatcat:jm7zjdgtxjarnnpcbpqj6xbsfy

Model for Trust Among Peers in Electronic Multiparty Transactions

Sanjay Goel, Jagdish Gangolly
2003 Americas Conference on Information Systems  
We propose a model that allows peers to form instant federations and collectively consummate complex business transactions. A transaction is done in two phases.  ...  We address the issue of trust among the principals in a multiparty transaction in a peer-to-peer system implementing a service-based architecture.  ...  In the first phase, the parties are passionate and peers are selected to form a federation based on the trust among the entities.  ... 
dblp:conf/amcis/GoelG03 fatcat:wxzedijdqjcadjybh5iigiubhy

On the advantages of P2P ML on mobile devices

Robert Basmadjian, Karim Boubouh, Amine Boussetta, Rachid Guerraoui, Alexandre Maurer
2022 Proceedings of the Thirteenth ACM International Conference on Future Energy Systems  
To motivate running ML algorithms on mobile devices, we also propose a new peer-to-peer personalized ML algorithm (P3) that shows better convergence properties than related works, and provably converging  ...  300 peers respectively.  ...  In the federated learning scheme, we rely on a server to aggregate the received updates from all the clients and use the aggregated models to learn a single global model.  ... 
doi:10.1145/3538637.3538863 fatcat:dachwrwmsrfo7hykhyrgzroque

Advances in privacy-preserving computing

Kaiping Xue, Zhe Liu, Haojin Zhu, Miao Pan, David S. L. Wei
2021 Peer-to-Peer Networking and Applications  
The twenty-second article by Longfei Zheng et al. on 'ASFGNN: Automated Separated-Federated Graph Neural Network' proposes an automated separated-federated Graph Neural Network (ASFGNN) learning paradigm  ...  Peer-to-Peer Netw. Appl. (2021) 14:1348-1352  ...  Wei focuses his research efforts on cloud and edge computing, IoT, 5G, big data, and machine learning.  ... 
doi:10.1007/s12083-021-01110-9 fatcat:o5vvf6ezcna2pc32g6oapioalu

Multigraph Topology Design for Cross-Silo Federated Learning [article]

Binh X. Nguyen, Tuong Do, Hien Nguyen, Vuong Pham, Toan Tran, Erman Tjiputra, Quang Tran, Anh Nguyen
2022 arXiv   pre-print
Cross-silo federated learning utilizes a few hundred reliable data silos with high-speed access links to jointly train a model.  ...  In this paper, we present a new multigraph topology for cross-silo federated learning. We first construct the multigraph using the overlay graph.  ...  Decentralized Federated Learning. Decentralized (or peer-to-peer) federated learning allows each silo data to interact with its neighbors directly without a central node [14] .  ... 
arXiv:2207.09657v2 fatcat:7t2karku2bhtdjrt6vnncbojla

Privacy-preserving Decentralized Aggregation for Federated Learning [article]

Beomyeol Jeon, S.M. Ferdous, Muntasir Raihan Rahman, Anwar Walid
2020 arXiv   pre-print
Federated learning is a promising framework for learning over decentralized data spanning multiple regions.  ...  Our secure aggregation protocol based on this novel group communication pattern design leads to an efficient algorithm for federated training with privacy guarantees.  ...  Similar to the classic federated learning, our algorithm consists of multiple rounds. In each FL round, every peer trains the model on its local dataset and updates the model parameters (U pdate).  ... 
arXiv:2012.07183v2 fatcat:fazprhigvzeb3algcza2soyf3i

Article Titles Index

2021 2021 7th International Conference on Signal Processing and Intelligent Systems (ICSPIS)  
Group Signature Based Federated Learning in Smart Grids 35. Automatic Epileptic Seizure Detection: Graph Features Versus Graph 56.  ...  Smart Grid based decentralized Peer-to-Peer Energy Trading Web Content Extraction by Weighing the Fundamental Contextual Rules 60.  ... 
doi:10.1109/icspis54653.2021.9729344 fatcat:qec5rksj4bctjilnrhr2f5dypa

A Self-Tuning Procedure for Resource Management in InterCloud Computing

Suleiman Onimisi Aliyu, Feng Chen, Han Li
2016 2016 IEEE International Conference on Software Quality, Reliability and Security Companion (QRS-C)  
to peer InterCloud.  ...  The focus of this paper is to propose a Software Cybernetic approach, in the form of an adaptive control framework, for efficient management of shared resources in peer-to-peer InterCloud computing.  ...  Based on architecture, Volunteer federations can further be classified into peer to peer or centralised.  ... 
doi:10.1109/qrs-c.2016.50 dblp:conf/qrs/AliyuCL16 fatcat:gbgo5gkzvfhw3gofic6nx7k7im

The database research group at the Max-Planck Institute for Informatics

Gerhard Weikum
2006 SIGMOD record  
In this setting, the peers' graph fragments may overlap arbitrarily, and peers are (a priori) unaware of other peers' fragments.  ...  Finding the top-k results involves an approximation to the Steiner tree problem, based on mimimum spanning trees on a precomputed connection graph, using a top-k threshold algorithm.  ... 
doi:10.1145/1168092.1168099 fatcat:zbcp54czabflhf5hn6jzz3nc4q

Trends and Advancements in Deep Neural Network Communication [article]

Felix Sattler, Thomas Wiegand, Wojciech Samek
2020 arXiv   pre-print
This paper gives an overview over the recent advancements and challenges in this new field of research at the intersection of machine learning and communications.  ...  Due to their great performance and scalability properties neural networks have become ubiquitous building blocks of many applications.  ...  In this section we will review the three most important settings, namely on-device inference, federated learning and peer-to-peer learning.  ... 
arXiv:2003.03320v1 fatcat:tgs7b6nbovflngyuxhxwkmxhv4

Implicit Model Specialization through DAG-based Decentralized Federated Learning [article]

Jossekin Beilharz, Bjarne Pfitzner, Robert Schmid, Paul Geppert, Bert Arnrich, Andreas Polze
2021 arXiv   pre-print
We propose a unified approach to decentralization and personalization in federated learning that is based on a directed acyclic graph (DAG) of model updates.  ...  Federated learning allows a group of distributed clients to train a common machine learning model on private data.  ...  ACKNOWLEDGMENTS We would like to thank the anonymous reviewers for their helpful comments on earlier versions of this paper.  ... 
arXiv:2111.01257v1 fatcat:fmcanel4drhldllvu4qe56wffy

Distributed databases and peer-to-peer databases

Angela Bonifati, Panos K. Chrysanthis, Aris M. Ouksel, Kai-Uwe Sattler
2008 SIGMOD record  
The need for large-scale data sharing between autonomous and possibly heterogeneous decentralized systems on the Web gave rise to the concept of P2P database systems.  ...  Whereas a definition for a P2P database system can be readily provided, a comparison with the more established decentralized models, commonly referred to as distributed, federated and multidatabases, is  ...  The authors would like to thank the anonymous reviewers for their valuable comments.  ... 
doi:10.1145/1374780.1374781 fatcat:som4flba6ne5nmwvkiosaetoji
« Previous Showing results 1 — 15 out of 22,389 results