3,746 Hits in 9.5 sec

General Proximal Incremental Aggregated Gradient Algorithms: Better and Novel Results under General Scheme [article]

Tao Sun, Yuejiao Sun, Dongsheng Li, Qing Liao
2019 arXiv   pre-print
Better and new convergence results are proved even with the general scheme.  ...  In this paper, we propose a general proximal incremental aggregated gradient algorithm, which contains various existing algorithms including the basic incremental aggregated gradient method.  ...  Conclusion In this paper, we consider a general proximal incremental aggregated gradient algorithm and prove several novel results. Much better results are proved under more general conditions.  ... 
arXiv:1910.05093v1 fatcat:ag4xb5tjabg4xhav62rxzvmjzm

PALMA, an improved algorithm for DOSY signal processing

Afef Cherni, Emilie Chouzenoux, Marc-André Delsuc
2017 The Analyst  
Acknowledgements This work was supported by the CNRS MASTODONS project under grant 2016TABASCO, by the Agence Nationale pour la Recherche, grant ANR2014 ONE SHOT 2D FT ICR.  ...  The authors thank Jean-Christophe Pesquet for the initial idea and for discussion all over this project.  ...  A new algorithmic approach based on a splitting scheme and on the use of proximity operators is introduced.  ... 
doi:10.1039/c6an01902a pmid:28120953 fatcat:x3mzmyq6vnaxhly7ozbevghhpm

Scalable Time-Decaying Adaptive Prediction Algorithm

Yinyan Tan, Zhe Fan, Guilin Li, Fangshan Wang, Zhengbing Li, Shikai Liu, Qiuling Pan, Eric P. Xing, Qirong Ho
2016 Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '16  
Under this observation, we thereby propose a novel time-decaying online learning algorithm derived from the state-of-the-art FTRL-proximal algorithm, called Time-Decaying Adaptive Prediction (TDAP) algorithm  ...  To scale Big Data, we further parallelize our algorithm following the data parallel scheme under both BSP and SSP consistency model.  ...  And better still, FTRL-proximal algorithm outperforms the other classical online learning algorithms (e.g., FOBOS and RDA) in terms of accuracy and sparsity, Therefore, in this paper, we derive our techniques  ... 
doi:10.1145/2939672.2939714 dblp:conf/kdd/TanFLWLLPXH16 fatcat:xmeo3bfzxjbrpei4u47sgrrwti

Federated Learning with Differential Privacy: Algorithms and Performance Analysis [article]

Kang Wei, Jun Li, Ming Ding, Chuan Ma, Howard H. Yang, Farokhi Farhad, Shi Jin, Tony Q. S. Quek, H. Vincent Poor
2019 arXiv   pre-print
Evaluations demonstrate that our theoretical results are consistent with simulations, thereby facilitating the designs on various privacy-preserving FL algorithms with different tradeoff requirements on  ...  Specifically, the theoretical bound reveals the following three key properties: 1) There is a tradeoff between the convergence performance and privacy protection levels, i.e., a better convergence performance  ...  The comparison of the loss function between experimental and theoretical results with the various aggregation times under NbAFL Algorithm with 50 clients.  ... 
arXiv:1911.00222v2 fatcat:z7suhf75obgqddyeilojzpydje

Nonasymptotic convergence of stochastic proximal point algorithms for constrained convex optimization [article]

Andrei Patrascu, Ion Necoara
2017 arXiv   pre-print
To avoid these drawbacks naturally introduced by the SGD scheme, the stochastic proximal point algorithms have been recently considered in the literature.  ...  For the newly introduced SPP scheme we prove new nonasymptotic convergence results.  ...  The research leading to these results has received funding from the Romanian National Authority for Scientific Research (UEFISCDI), PNII- RU-TE 2014, project MoCOBiDS, no. 176/01.10.2015  ... 
arXiv:1706.06297v1 fatcat:npkuttlnvje7bduedvpz6rbude

Geometric algorithms for sensor networks

J. Gao, L. Guibas
2011 Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences  
The physical locations of the sensor nodes greatly impact on system design in all aspects, from low-level networking and organization to high-level information processing and applications.  ...  This paper reviews work in the past 10 years on topics such as network localization, geometric routing, information discovery, data-centric routing and topology discovery.  ...  Generally speaking, localization algorithms can be classified as anchor-based and anchor-free methods.  ... 
doi:10.1098/rsta.2011.0215 pmid:22124080 fatcat:2m3bwwdjnramzcrtkjip6om3t4

Gradient-Consensus: Linearly Convergent Distributed Optimization Algorithm over Directed Graphs [article]

Vivek Khatana, Govind Saraswat, Sourav Patel, Murti V. Salapaka
2021 arXiv   pre-print
The communication overhead for the improved guarantees on meeting constraints and better convergence of our algorithm is O(klog k) iterates in comparison to O(k) of the traditional algorithms.  ...  We establish that the proposed algorithm has a global R-linear rate of convergence if the aggregate function f is strongly convex and Lipschitz differentiable.  ...  The works in [12] , [13] have developed algorithms based on proximal-gradient to tackle (1) with proximal friendly f i 's.  ... 
arXiv:1909.10070v7 fatcat:uodxxuvykzdepkzrwhr5rbst5a

Optimization Algorithms as Robust Feedback Controllers [article]

Adrian Hauswirth, Saverio Bolognani, Gabriela Hug, Florian Dörfler
2021 arXiv   pre-print
However, our focus lies on recent methods under the name of "feedback-based optimization".  ...  While this new perspective is insightful in itself, liberating optimization methods from specific numerical and algorithmic aspects opens up new possibilities to endow complex real-world systems with sophisticated  ...  Acknowledgements The research leading to these results was supported by ETH Zürich funds and by the Swiss Federal Office of Energy grant #SI/501708 UNICORN.  ... 
arXiv:2103.11329v1 fatcat:awpupmsxirdw5lnbsvmkhrs57a

A Distributed Flexible Delay-tolerant Proximal Gradient Algorithm [article]

Konstantin Mishchenko, Franck Iutzeler, Jérôme Malick
2019 arXiv   pre-print
The obtained rates are the same as the vanilla proximal gradient algorithm over some introduced epoch sequence that subsumes the delays of the system.  ...  We develop and analyze an asynchronous algorithm for distributed convex optimization when the objective writes a sum of smooth functions, local to each worker, and a non-smooth function.  ...  It is closely related to the proximal incremental aggregated gradient (PIAG) method [1, 28] .  ... 
arXiv:1806.09429v3 fatcat:wxlwb6hitjek5bdcwlwdqup5gi

FedDR – Randomized Douglas-Rachford Splitting Algorithms for Nonconvex Federated Composite Optimization [article]

Quoc Tran-Dinh, Nhan H. Pham, Dzung T. Phan, Lam M. Nguyen
2021 arXiv   pre-print
Our algorithms rely on a novel combination between a nonconvex Douglas-Rachford splitting method, randomized block-coordinate strategies, and asynchronous implementation.  ...  In fact, our new algorithms match the communication complexity lower bound up to a constant factor under standard assumptions.  ...  The authors would also like to thank all the anonymous reviewers and the ACs for their constructive comments to improve the paper.  ... 
arXiv:2103.03452v3 fatcat:pc4lnvomqngzbb5du5kf3mnwlq

A novel approach for initializing the spherical K-means clustering algorithm

Rehab Duwairi, Mohammed Abu-Rahmeh
2015 Simulation modelling practice and theory  
Starting with such seeds gives the opportunity for the algorithm to find better clusters, and faster convergence conditions.  ...  Normally the proximity measure will be the cosine similarity, when combined with the K-means, the resulting algorithm is known as the spherical K-means.  ...  He verified the supremacy of using a gradually decreasing learning rate to the flat rate, relating the incremental mode to the gradient ascent approach.  ... 
doi:10.1016/j.simpat.2015.03.007 fatcat:hwgplyqftraubk7w7bdsdidp7y

Paradigms for Algorithms and Interactions [chapter]

Andrea Zanella, Michele Zorzi, Elena Fasolo, Anibal Ollero, Ivan Maza, Antidio Viguria, Marcelo Pias, George Coulouri, Chiara Petrioli
2010 Cooperating Embedded Systems and Wireless Sensor Networks  
Embedded WiSeNts Paradigms for algorithms and interactions  ...  This document provides a survey of the most important design paradigms, algorithms and interaction patterns that characterize the systems based on Cooperating Objects.  ...  If there are overlapping paths, they are combined to form aggregation tree. • Greedy incremental tree: this is a sequential scheme.  ... 
doi:10.1002/9780470610817.ch3 fatcat:y6rg4ciutvelxkwt72fhwzdyd4

LoCUS: A multi-robot loss-tolerant algorithm for surveying volcanic plumes [article]

John Erickson, Abhinav Aggarwal, G. Matthew Fricke, Melanie E. Moses
2020 arXiv   pre-print
Further, the novel data-structures and algorithms underpinning LoCUS have application in other areas of fault-tolerant algorithm research.  ...  Drones must be able to follow gas concentration gradients while tolerating frequent drone loss. We present the LoCUS algorithm as a solution to this problem and prove its robustness.  ...  [11] compares 3 algorithms for plume gradient following: the surge-cast algorithm, the Dung Beetle (zig-zag) algorithm, and the pseudo-gradient algorithm using a single robot agent.  ... 
arXiv:2009.00156v1 fatcat:yefaxwofyjeffm23km4fu32wqq

A Comparison of Self-Play Algorithms Under a Generalized Framework [article]

Daniel Hernandez, Kevin Denamganai, Sam Devlin, Spyridon Samothrakis, James Alfred Walker
2020 arXiv   pre-print
They allow to verify and replicate existing findings, and to link is connected results.  ...  Our results indicate that, throughout training, various self-play definitions exhibit cyclic policy evolutions.  ...  Gupta for his insightful conversations and work on Nash averaging.  ... 
arXiv:2006.04471v1 fatcat:6jpsg2hpknb7jbkrd7pmhrbn5u

Monkey Behavior Based Algorithms - A Survey

R. Vasundhara Devi, S. Siva Sathya
2017 International Journal of Intelligent Systems and Applications  
In this survey, we provide a comprehensive overview of monkey behavior based algorithms and their related literatures and discuss useful research directions to provide better insights for swarm intelligence  ...  Since then, several variants such as Monkey search, Monkey algorithm, and Spider Monkey optimization algorithms have been proposed.  ...  SMO is considered to handle the premature convergence and stagnation efficiently and could result in better solution.  There is a global leader who is the oldest female member in the group.  ... 
doi:10.5815/ijisa.2017.12.07 fatcat:me7s6qz4dbhxlmr7ltswxelguu
« Previous Showing results 1 — 15 out of 3,746 results