A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2024; you can also visit the original URL.
The file type is application/pdf
.
Filters
opML: Optimistic Machine Learning on Blockchain
[article]
2024
arXiv
pre-print
The integration of machine learning with blockchain technology has witnessed increasing interest, driven by the vision of decentralized, secure, and transparent AI services. ...
In this context, we introduce opML (Optimistic Machine Learning on chain), an innovative approach that empowers blockchain systems to conduct AI model inference. opML lies a interactive fraud proof protocol ...
DNN Computation in Multi-Phase opML In this demonstration, we present a DNN computation in a two-phase opML approach: • The computation process of Machine Learning, specifically Deep Neural Networks (DNN ...
arXiv:2401.17555v2
fatcat:mjnor3j3sbbz3kc4qfqipthstm
Distributed Deep Learning: From Single-Node to Multi-Node Architecture
2022
Electronics
Local parallelism is considered quite important in the design of a time-performing multi-node architecture because DDL depends on the time required by all the nodes. ...
During the last years, deep learning (DL) models have been used in several applications with large datasets and complex models. ...
The problem is that by doing this, the network communication is intensified. For example when the neural network is divided [28] on several machines. ...
doi:10.3390/electronics11101525
fatcat:eemwzpv4ifh4xlgvbx7hknsu5y
In-database distributed machine learning
2019
Proceedings of the VLDB Endowment
In this demonstration, we give a practical exhibition of a solution for the enablement of distributed machine learning natively inside database engines. ...
Machine learning has enabled many interesting applications and is extensively being used in big data systems. ...
Audience will experience the entire machine learning workflow presented in Figure 3 . The audience will be free to vary the number of iterations or the number of nodes in neural network model. ...
doi:10.14778/3352063.3352083
fatcat:uz7cagmlpzcg5mbmxt75xurguu
Energy-based Graph Convolutional Networks for Scoring Protein Docking Models
[article]
2019
bioRxiv
pre-print
Directly learning from structure data in graph representation, EGCN represents the first successful development of graph convolutional networks for protein docking. ...
In this study the two challenging problems in protein docking are regarded as relative and absolute scoring, respectively, and addressed in one physics-inspired deep learning framework. ...
ACKNOWLEDGEMENTS This work was supported by the National Institutes of Health (R35GM124952). ...
doi:10.1101/2019.12.19.883371
fatcat:5t2qjbyaczhtthjsvdph4tjwue
Analysis of Intrusion Detection and Classification using Machine Learning Approaches
2017
INTERNATIONAL JOURNAL ONLINE OF SCIENCE
This paper discusses some usually used machine learning techniques in Intrusion Detection System and conjointly reviews a number of the prevailing machine learning IDS proposed by researchers at different ...
Analysis showed that application of machine learning techniques in intrusion detection might reach high detection rate. ...
[12] in like manner connected a multi-level model with various machine learning procedures, for example, C5, MLP, and Naïve Bayes. ...
doi:10.24113/ijoscience.v3i10.13
fatcat:lp56zcmjlzduxipfvyqhskuhky
Snap ML: A Hierarchical Framework for Machine Learning
[article]
2018
arXiv
pre-print
The framework, named Snap Machine Learning (Snap ML), combines recent advances in machine learning systems and algorithms in a nested manner to reflect the hierarchical architecture of modern computing ...
We evaluate the performance of Snap ML in both single-node and multi-node environments, quantifying the benefit of the hierarchical scheme and the data streaming functionality, and comparing with other ...
*Trademark, service mark, registered trademark of International Business Machines Corporation in the United States, other countries, or both. ** Intel Xeon is a trademarks or registered trademarks of Intel ...
arXiv:1803.06333v3
fatcat:l75n5irwuzcatgwrw5jhgvxro4
Energy-based Graph Convolutional Networks for Scoring Protein Docking Models
[article]
2019
arXiv
pre-print
Directly learning from 3D structure data in graph representation, EGCN represents the first successful development of graph convolutional networks for protein docking. ...
In this study the two challenging problems in protein docking are regarded as relative and absolute scoring, respectively, and addressed in one physics-inspired deep learning framework. ...
ACKNOWLEDGEMENTS This work was supported by the National Institutes of Health (R35GM124952). ...
arXiv:1912.12476v1
fatcat:v5yguzbbi5dajo5juh6hjouowu
Navigating the maze of graph analytics frameworks using massive graph datasets
2014
Proceedings of the 2014 ACM SIGMOD international conference on Management of data - SIGMOD '14
Implementing graph traversal, statistics and machine learning algorithms on such data in a scalable manner is quite challenging. ...
In this work, we offer a quantitative roadmap for improving the performance of all these frameworks and bridging the "ninja gap". ...
The native code uses MPI message passing [6] to drive the underlying fabric (FDR InfiniBand network in our case) for high bandwidth and low latency communication among the nodes. ...
doi:10.1145/2588555.2610518
dblp:conf/sigmod/SatishSPSPHSYD14
fatcat:l5mcilp3ujezbliklr4h2pwrfe
MPI applications' performances in native vs. virtualized environments using InfiniBand IPoIB virtualization and live migration
2015
Tehnički Vjesnik
This paper presents a face-to-face native-to-virtual HPC MPI application performance analysis on a range of general-use HPC applications by virtualizing InfiniBand via IPoIB at the guest virtual machine ...
architecture on a proactive fault-tolerance use case. ...
The guest operating system in the virtual machines, running on top of KVM and ESXi was configured within the same environment as the native nodes' using Puppet configuration management. ...
doi:10.17559/tv-20140819115232
fatcat:uzhp7s2xtff4lnembrqcofbwyq
Hyper: Distributed Cloud Processing for Large-Scale Deep Learning Tasks
[article]
2019
arXiv
pre-print
Training and deploying deep learning models in real-world applications require processing large amounts of data. ...
The system implements a distributed file system and failure-tolerant task processing scheduler, independent of the language and Deep Learning framework used. ...
Also would like to thank AWS for providing cloud resources for experiments. The project was funded and supported by Snark AI, Inc. ...
arXiv:1910.07172v1
fatcat:5qlpul5yqzfyrhfgallvnyj4ne
Towards Quantum-Enabled 6G Slicing
[article]
2022
arXiv
pre-print
The quantum machine learning (QML) paradigms and their synergies with network slicing can be envisioned to be a disruptive technology on the cusp of entering to era of sixth-generation (6G), where the ...
In this intent, we propose a cloud-native federated learning framework based on quantum deep reinforcement learning (QDRL) where distributed decision agents deployed as micro-services at the edge and cloud ...
The decision agents are deployed in edge nodes while federation layers at the cloud node in our Kubernetes infrastructure connected to 5G segments. ...
arXiv:2212.11755v1
fatcat:kth3nhn245fjdjia5ngk7nwhje
A Data-Centric Optimization Framework for Machine Learning
[article]
2022
arXiv
pre-print
Rapid progress in deep learning is leading to a diverse set of quickly changing models, with a dramatically growing demand for compute. ...
The pipeline begins with standard networks in PyTorch or ONNX and transforms computation through progressive lowering. ...
Guided Optimization Case Study: EfficientNet In this case study, we consider the EfficientNet-B0 [57] of the network. ...
arXiv:2110.10802v3
fatcat:obpwqoanuzdcdf7j7mgh33fzdu
Secure Self Optimizing Software Defined Framework for NB-IoT Towards 5G
2020
Procedia Computer Science
In this article self-optimizing framework for ultra IoT is proposed, which represents the M2M communication, network cloud. ...
In this article self-optimizing framework for ultra IoT is proposed, which represents the M2M communication, network cloud. ...
To enhance the customer experience machine learning cloud native solution with reduction in cost and congestion, ultra-traffic optimization heuristic optimizes the traffic flows on congested cells and ...
doi:10.1016/j.procs.2020.04.298
fatcat:urm7c2i7dbectjeodzkjsfulg4
First Scalable Machine Learning Based Architecture for Cloud-native Transport SDN Controller
2021
Zenodo
We present a cloud-native architecture with a machine learning QoT predictor that enables cognitive functions in transport SDN controllers. ...
We evaluate the QoT predictor training and auto-scaling capabilities in a real WDM/SDM testbed. ...
Acknowledgments Work supported by the EC H2020 TeraFlow (101015857) and Spanish AURORAS (RTI2018-099178-I00) and Ministry of Internal Affairs and Communications, Japan grant number JP MI00316. ...
doi:10.5281/zenodo.5087536
fatcat:63yq3yjearcw3gfxhwej5bsor4
A Scalable and Cloud-Native Hyperparameter Tuning System
[article]
2020
arXiv
pre-print
In this paper, we introduce Katib: a scalable, cloud-native, and production-ready hyperparameter tuning system that is agnostic of the underlying machine learning framework. ...
We present the motivation and design of the system and contrast it with existing hyperparameter tuning systems, especially in terms of multi-tenancy, scalability, fault-tolerance, and extensibility. ...
., number of clusters in k-means clustering, learning rate, batch size, and number of hidden nodes in neural networks) cannot be learnt during the training process, unlike the value of model parameters ...
arXiv:2006.02085v2
fatcat:msydmwpvzne4rlhjvmc5sxiike
« Previous
Showing results 1 — 15 out of 23,310 results