193 Hits in 7.1 sec

FLeet: Online Federated Learning via Staleness Awareness and Performance Prediction [article]

Georgios Damaskinos, Rachid Guerraoui, Anne-Marie Kermarrec, Vlad Nitu, Rhicheek Patra, Francois Taiani
2020 arXiv   pre-print
FLeet combines the privacy of Standard FL with the precision of online learning thanks to two core components: (i) I-Prof, a new lightweight profiler that predicts and controls the impact of learning tasks  ...  This paper presents FLeet, the first Online FL system, acting as a middleware between the Android OS and the machine learning application.  ...  FLeet also makes use of AdaSGD, a new staleness-aware learning algorithm that is optimized for Online FL.  ... 
arXiv:2006.07273v1 fatcat:qfic5skl6bhxfjq5ses2wedaxq

Programming and Deployment of Autonomous Swarms using Multi-Agent Reinforcement Learning [article]

Jayson Boubin, Codi Burley, Peida Han, Bowen Li, Barry Porter, Christopher Stewart
2021 arXiv   pre-print
To conserve compute resources, the Fleet Computer gives priority scheduling to models that contribute to effective actions, drawing a novel link between online learning and resource management.  ...  Using just two programmer-provided functions Map() and Eval(), the Fleet Computer compiles and deploys swarms and continuously updates the reinforcement learning models that govern actions.  ...  Figure 5 (a-b) shows the Fleet Computer's performance on our crop scouting and video analytics workloads without additional online learning.  ... 
arXiv:2105.10605v1 fatcat:xddio3i75bg73ng4p2u4vqirei

Addressing modern and practical challenges in machine learning: A survey of online federated and transfer learning [article]

Shuang Dai, Fanlin Meng
2022 arXiv   pre-print
Online federated learning (OFL) and online transfer learning (OTL) are two collaborative paradigms for overcoming modern machine learning challenges such as data silos, streaming data, and data security  ...  This survey explored OFL and OTL throughout their major evolutionary routes to enhance understanding of online federated and transfer learning.  ...  The latter is a novel stochastic gradient descent algorithm that employs weighted stale gradients determined by the stale-aware dampening factor and the similarity-based boosting value.  ... 
arXiv:2202.03070v1 fatcat:aiuhmm5dejbhxpn4nkxutpltry

Resource-Efficient Federated Learning [article]

Ahmed M. Abdelmoniem and Atal Narayan Sahu and Marco Canini and Suhaib A. Fahmy
2021 arXiv   pre-print
Federated Learning (FL) enables distributed training by learners using local data, thereby enhancing privacy and reducing communication.  ...  and bias.  ...  Training models using this approach is known as Federated Learning (FL).  ... 
arXiv:2111.01108v1 fatcat:r6iuki3cpnfhlo7rr5fzhilhkq

A Taxonomy of Software Engineering Challenges for Machine Learning Systems: An Empirical Investigation [chapter]

Lucy Ellen Lwakatare, Aiswarya Raj, Jan Bosch, Helena Holmström Olsson, Ivica Crnkovic
2019 Lecture Notes in Business Information Processing  
In federated learning an initial model is built locally and then it gets trained and improved at the edge.  ...  It exposes a REST API that can be called via JavaScript fetch from the online music catalogue application.  ... 
doi:10.1007/978-3-030-19034-7_14 fatcat:6wy5wa3l6zcy3ehq5axm77xabm

Advances and Open Problems in Federated Learning [article]

Peter Kairouz, H. Brendan McMahan, Brendan Avent, Aurélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista Bonawitz, Zachary Charles, Graham Cormode, Rachel Cummings, Rafael G.L. D'Oliveira, Hubert Eichner (+47 others)
2021 arXiv   pre-print
Federated learning (FL) is a machine learning setting where many clients (e.g. mobile devices or whole organizations) collaboratively train a model under the orchestration of a central server (e.g. service  ...  FL embodies the principles of focused data collection and minimization, and can mitigate many of the systemic privacy risks and costs resulting from traditional, centralized machine learning and data science  ...  Acknowledgments The authors would like to thank Alex Ingerman and David Petrou for their useful suggestions and insightful comments during the review process.  ... 
arXiv:1912.04977v3 fatcat:efkbqh4lwfacfeuxpe5pp7mk6a

D1.1 - State of the Art Analysis

Danilo Ardagna
2021 Zenodo  
Then, the deliverable provides a background on AI applications design, also considering some advanced design trends (e.g., Network Architecture Search, Federated Learning, Deep Neural Networks partitioning  ...  ), providing resource efficiency, performance, data privacy, and security guarantees.  ...  It offers: • High-performance online FL algorithms • Dynamic loading federated learning models • Can serve multiple models, or multiple versions of the same model • Real-time inference • Support pre-and  ... 
doi:10.5281/zenodo.6372377 fatcat:f6ldfuwivbcltew4smiiwphfty


2021 2021 IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)  
as online learning activity.  ...  and Prediction using Decision Tree for SPOC Teaching Chunyan Yu and Qi Hui (Chuzhou University, China) SPOC(Small Private Online courses) is a blended learning model using MOOC  ... 
doi:10.1109/icce-tw52618.2021.9602919 fatcat:aetmvxb7hfah7iuucbamos2wgu

D4.3 – WP4 Scientific Report and Prototype Description – Y3

Yosef Moatti, Paula Ta Shma, Guy Khazma, Javier López Moratalla, Jacob Roldan, Rogelio Rodriguez, Luis Tomás Bolívar, Marta Patiño, Ainhoa Azqueta, George Makridis, Christos Doulkeridis, Maria Kanakari (+4 others)
2021 Zenodo  
An initial demonstration of the capabilities offered by the data services has been performed during the interim review of the project, in which all the components have been integrated and interacted to  ...  The data services of this environment are naturally at the core of BigDataStack and are covered in this deliverable in terms of design specification as well as in terms of integration and experimentation  ...  To discover possible deviations in data, we need to learn and predict the relationships between data points.  ... 
doi:10.5281/zenodo.4442344 fatcat:dd3d7d3hofcp5gn7lrmxwmrdr4

Digitally Connected: Global Perspectives on Youth and Digital Media

Sandra Cortesi, Urs Gasser, Gameli Adzaho, Bruce Baikie, Jacqueline Baljeu, Matthew Battles, Jacqueline Beauchere, Elsa Brown, Jane O Burns, Patrick Burton, Jasmina Byrne, Maximillion Colombo (+37 others)
2015 Social Science Research Network  
The opportunity to share our experiences with our international colleagues was invaluable, and we eagerly look forward to a collaborative future. References / Resources / Links  ...  "information about an actor's past performance that helps predict the actor's future ability to perform or to satisfy the decision-maker's preferences" (Goldman, 2010, p.294 ).  ...  Young people operate in the online space across every facet of their life, working, learning, socialising and playing via the Internet.  ... 
doi:10.2139/ssrn.2585686 fatcat:ehi5kz4qmrc2ni5kiqfc5btzbe

Educating the Educators

Tina M. Adams, R. Sean Evans
2004 Journal of Library Administration  
of Education regarding provision and promotion of Library services to students and faculty taking and teaching classes virtually and at satellite campus and computer lab locations throughout the state  ...  The focus of "Educating the Educators: Outreach to the College of Education Distance Faculty and Native American Students" is to examine and explore the relationship between the Library and the College  ...  We can use the videoconferencing equipment as a way to communicate, teach and learn from the School of Human Services staff.  ... 
doi:10.1300/j111v41n01_02 fatcat:nwphvdzmmzc4rdr7ghat52itky

Distributed Deep Learning in Open Collaborations [article]

Michael Diskin, Alexey Bukhtiyarov, Max Ryabinin, Lucile Saulnier, Quentin Lhoest, Anton Sinitsin, Dmitry Popov, Dmitry Pyrkin, Maxim Kashirin, Alexander Borzunov, Albert Villanova del Moral, Denis Mazur (+4 others)
2021 arXiv   pre-print
To address this demand, large corporations and institutions use dedicated High-Performance Computing clusters, whose construction and maintenance are both environmentally costly and well beyond the budget  ...  We demonstrate the effectiveness of our approach for SwAV and ALBERT pretraining in realistic conditions and achieve performance comparable to traditional setups at a fraction of the cost.  ...  Also, we thank Abhishek Thakur for helping with downstream evaluation and Tanmoy Sarkar with Omar Sanseviero, who helped us organize the collaborative experiment and gave regular status updates to the  ... 
arXiv:2106.10207v2 fatcat:3rf2r2aerzf77dsw34e4c2psbm

Artificial Intellgence – Application in Life Sciences and Beyond. The Upper Rhine Artificial Intelligence Symposium UR-AI 2021 [article]

Karl-Herbert Schäfer
2021 arXiv   pre-print
) and the University of Applied Sciences and Arts Northwestern Switzerland.  ...  Topics of the conference are applications of Artificial Intellgence in life sciences, intelligent systems, industry 4.0, mobility and others.  ...  (Charité, Institute of Pathology) for helpful comments and acknowledge the financial support by the Federal Ministry of Education and Research of Germany (BMBF) in the project deep.HEALTH (13FH770IX6).  ... 
arXiv:2112.05657v1 fatcat:wdjgymicyrfybg5zth2dc2i3ni

Institutionalized Word Taboo: The Continuing Saga of FCC Indecency Regulation

Christopher M. Fairman
2013 Social Science Research Network  
Social cognitive theory distinguishes between acquisition and performance because people do not perform everything they learn. 365 Whether "observers actually engage in that learned behavior is a function  ...  T's UVerse, Verizon's FIOS, and DirecTV; over the Internet on popular websites such as YouTube, iTunes, and Hulu; via podcasts; by online video streaming through services such as Netflix; and through DVD  ... 
doi:10.2139/ssrn.2223992 fatcat:26dpxc4yvjd5xlepx7nuey6ica

Chasing the AIDS virus

Thomas Lengauer, André Altmann, Alexander Thielen, Rolf Kaiser
2010 Communications of the ACM  
I also thank Tuomas Sandholm for feedback on the kidney exchange section, and David Pennock for feedback on the prediction markets section.  ...  The authors find that, surprisingly, their filter-based gist is rather good at predicting the number of instances of a given object category that might be present in the scene, as well as their likely  ...  to perform energy-aware scheduling.  ... 
doi:10.1145/1666420.1666440 fatcat:o2qllqh4tzhh5dzgvnjewl52vq
« Previous Showing results 1 — 15 out of 193 results