Exploring the computational cost of machine learning at the edge for human-centric Internet of Things

Gómez-Carmona, Oihane; Casado-Mansilla, Diego; Kraemer, Frank Alexander; López-de-Ipiña, Diego; García-Zubía, Javier

doi:10.1016/j.future.2020.06.013

Cited by 32 publications

(16 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Between the factors that affect the computational cost, there is the number of features extracted from the data. Gómez-Carmona et al [ 48 ], for example, claim that computational effort can be reduced by 80% assuming a decline of the classification accuracy of only 3%. This work utilizes data pre-processing, a three-point median filter to smooth the signal followed by a segmentation process and a feature selection method (Chi2 filtering [ 88 ]), which consists of a discriminating process to find essential features that have more weight in the model.…”

Section: Answering the Rqsmentioning

confidence: 99%

See 1 more Smart Citation

A Systematic Literature Review on Distributed Machine Learning in Edge Computing

Filho

Marques

Chang

et al. 2022

Sensors

View full text Add to dashboard Cite

Distributed edge intelligence is a disruptive research area that enables the execution of machine learning and deep learning (ML/DL) algorithms close to where data are generated. Since edge devices are more limited and heterogeneous than typical cloud devices, many hindrances have to be overcome to fully extract the potential benefits of such an approach (such as data-in-motion analytics). In this paper, we investigate the challenges of running ML/DL on edge devices in a distributed way, paying special attention to how techniques are adapted or designed to execute on these restricted devices. The techniques under discussion pervade the processes of caching, training, inference, and offloading on edge devices. We also explore the benefits and drawbacks of these strategies.

show abstract

Section: Answering the Rqsmentioning

confidence: 99%

“…On the contrary, it requires even more concern, as we are talking about extremely sensitive data. In response to the demand for privacy, trust and control over the data, executing machine learning tasks at the edge of the system has the potential to make the healthcare services more human-centric [ 48 ].…”

Section: Answering the Rqsmentioning

confidence: 99%

A Systematic Literature Review on Distributed Machine Learning in Edge Computing

Filho

Marques

Chang

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Likewise, the growing interest in data privacy safeguarding is reflected in emerging legal frames such as the General Data Protection Regulation (GDPR) [3]. The second challenge is related to the increasing availability of data, which, on the one hand, is furthering the progress of artificial intelligence [4], and, on the other hand, it arises new challenges related to its storage and processing that are even exacerbated when data stemmed from distributed sources, as in IoT scenarios [5]. The latter challenge emerges from the need to distributively process data when it is not possible to transfer it to a central server, because of legal or regulatory restrictions, communication costs or other kind of technical limitations.…”

Section: Introductionmentioning

confidence: 99%

Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challenges

Rodríguez-Barroso¹,

López²,

Luzón³

et al. 2022

Preprint

View full text Add to dashboard Cite

Federated learning is a machine learning paradigm that emerges as a solution to the privacy-preservation demands in artificial intelligence. As machine learning, federated learning is threatened by adversarial attacks against the integrity of the learning model and the privacy of data via a distributed approach to tackle local and global learning. This weak point is exacerbated by the inaccessibility of data in federated learning, which makes harder the protection against adversarial attacks and evidences the need to furtherance the research on defence methods to make federated learning a real solution for safeguarding data privacy. In this paper, we present an extensive review of the threats of federated learning, as well as as their corresponding countermeasures, attacks versus defences. This survey provides a taxonomy of adversarial attacks and a taxonomy of defence methods that depict a general picture of this vulnerability of federated learning and how to overcome it. Likewise, we expound guidelines for selecting the most adequate defence method according to the category of the adversarial attack. Besides, we carry out an extensive experimental study from which we draw further conclusions about the behaviour of attacks and defences and the guidelines for selecting the most adequate defence method according to the category of the adversarial attack. This study is finished leading to meditated learned lessons and challenges.

show abstract

“…From the perspective of utilizing edge computing for increasing the efficiency of IoT, several works have proposed and evaluated ML techniques for higher energy efficiency, bandwidth saving, lower latency, and collaborative intelligence of the network [ 1 , 10 , 11 , 12 , 13 , 14 , 15 ]. However, most of these works provide analytical models with simulation-based results that cannot be entirely relied upon for the real deployed networks because simulation-based validations do not accurately portray the empirical measurements of real-life systems.…”

Section: Introductionmentioning

confidence: 99%

An NB-IoT-Based Edge-of-Things Framework for Energy-Efficient Image Transfer

Khan

Moullec

Alam

2021

Sensors

View full text Add to dashboard Cite

Machine Learning (ML) techniques can play a pivotal role in energy efficient IoT networks by reducing the unnecessary data from transmission. With such an aim, this work combines a low-power, yet computationally capable processing unit, with an NB-IoT radio into a smart gateway that can run ML algorithms to smart transmit visual data over the NB-IoT network. The proposed smart gateway utilizes supervised and unsupervised ML algorithms to optimize the visual data in terms of their size and quality before being transmitted over the air. This relaxes the channel occupancy from an individual NB-IoT radio, reduces its energy consumption and also minimizes the transmission time of data. Our on-field results indicate up to 93% reductions in the number of NB-IoT radio transmissions, up to 90.5% reductions in the NB-IoT radio energy consumption and up to 90% reductions in the data transmission time.

show abstract

Exploring the computational cost of machine learning at the edge for human-centric Internet of Things

Cited by 32 publications

References 64 publications

A Systematic Literature Review on Distributed Machine Learning in Edge Computing

A Systematic Literature Review on Distributed Machine Learning in Edge Computing

Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challenges

An NB-IoT-Based Edge-of-Things Framework for Energy-Efficient Image Transfer

Contact Info

Product

Resources

About