A Comparative Study of Classification Techniques for Managing IoT Devices of Common Specifications

Mavrogiorgou, Argyro; Kiourtis, Athanasios; Kyriazis, Dimosthenis

doi:10.1007/978-3-319-68066-8_6

Cited by 15 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This minor drop in the accuracy may be due to the fact that NNs are more likely to overfit and can suffer from multiple local minima compared to SVMs. Overall, the true negative classification remains high in all cases and only the false positive rate shows a variation, depending on the internal way of operation of each classifier [19].…”

Section: Results and Analysismentioning

confidence: 92%

Unsupervised Learning for Trustworthy IoT

Banerjee

Giannetsos

Panaousis

et al. 2018

2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

View full text Add to dashboard Cite

The advancement of Internet-of-Things (IoT) edge devices with various types of sensors enables us to harness diverse information with Mobile Crowd-Sensing applications (MCS). This highly dynamic setting entails the collection of ubiquitous data traces, originating from sensors carried by people, introducing new information security challenges; one of them being the preservation of data trustworthiness. What is needed in these settings is the timely analysis of these large datasets to produce accurate insights on the correctness of user reports. Existing data mining and other artificial intelligence methods are the most popular to gain hidden insights from IoT data, albeit with many challenges. In this paper, we first model the cyber trustworthiness of MCS reports in the presence of intelligent and colluding adversaries. We then rigorously assess, using real IoT datasets, the effectiveness and accuracy of well-known data mining algorithms when employed towards IoT security and privacy. By taking into account the spatiotemporal changes of the underlying phenomena, we demonstrate how concept drifts can masquerade the existence of attackers and their impact on the accuracy of both the clustering and classification processes. Our initial set of results clearly show that these unsupervised learning algorithms are prone to adversarial infection, thus, magnifying the need for further research in the field by leveraging a mix of advanced machine learning models and mathematical optimization techniques.

show abstract

Section: Results and Analysismentioning

confidence: 92%

Unsupervised Learning for Trustworthy IoT

Banerjee

Giannetsos

Panaousis

et al. 2018

2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

View full text Add to dashboard Cite

show abstract

“…IoT devices: IoT devices can be any devices (e.g., sensor or actuators) [31], and their detail specification varies depending on the actual use cases. For example, devices in a "smart forest or smart agriculture" environment, which are most likely to reside in rural areas, may have lower computing, storage, and networking resources than those in the "smart factory or smart home" sector.…”

Section: Iot Domainsmentioning

confidence: 99%

A Conceptual Architecture in Decentralizing Computing, Storage, and Networking Aspect of IoT Infrastructure

Oktian

Witanto

Lee

2021

IoT

View full text Add to dashboard Cite

Since the inception of the Internet of Things (IoT), we have adopted centralized architecture for decades. With the vastly growing number of IoT devices and gateways, this architecture struggles to cope with the high demands of state-of-the-art IoT services, which require scalable and responsive infrastructure. In response, decentralization becomes a considerable interest among IoT adopters. Following a similar trajectory, this paper introduces an IoT architecture re-work that enables three spheres of IoT workflows (i.e., computing, storage, and networking) to be run in a distributed manner. In particular, we employ the blockchain and smart contract to provide a secure computing platform. The distributed storage network maintains the saving of IoT raw data and application data. The software-defined networking (SDN) controllers and SDN switches exist in the architecture to provide connectivity across multiple IoT domains. We envision all of those services in the form of separate yet integrated peer-to-peer (P2P) overlay networks, which IoT actors such as IoT domain owners, IoT users, Internet Service Provider (ISP), and government can cultivate. We also present several IoT workflow examples showing how IoT developers can adapt to this new proposed architecture. Based on the presented workflows, the IoT computing can be performed in a trusted and privacy-preserving manner, the IoT storage can be made robust and verifiable, and finally, we can react to the network events automatically and quickly. Our discussions in this paper can be beneficial for many people ranging from academia, industries, and investors that are interested in the future of IoT in general.

show abstract

“…The author in [17] discussed four classification algorithms that are used, such as KNN, NB, SVM, RF are applied to a data set containing the specifications of known devices for classification. They compare the classifiers using performance metrics and shows that the KNN is the best classifier.…”

Section: Iot Devices Classification Based On Machine Learningmentioning

confidence: 99%

“…We have reviewed the paper for classifying the IoT devices using various machine learning classifiers shows in Fig 3. The author uses RF [15,16,17,19,21], Bagging [15], NB [17,19], DT [21], SVM [17,21], KNN [17,21], ANN [21], CNN [21], MLP [18], LSTM [18], and LSTM-CNN [18] classifiers. Fig.…”

Section: Iot Devices Classification Based On Machine Learningmentioning

confidence: 99%

Device Classification for Machine to Machine Communication in Internet of Things for Tele-Robotic Surgery: A Review

Lokhande¹,

Patil²

2021

INDJCSE

View full text Add to dashboard Cite

Internet of Things (IoT) domain gradually growing in information technology and telecommunications systems area considering its size and complexity. It is recognized as potential domain in IoT applications. New technologies need to anticipate this sharp increase in the number of varying smart devices. In current perspective, the IOT environment, which allows a huge network of things to communicate with each other, may face a number of technical and application problems, such as privacy concerns, security, sensor anonymity, decision support, a variety of device applications, and so on. In this review paper, goal is to study various frameworks for IoT device classification possible in Tele robotic surgery and summarizes the classification algorithm. Tele robotic surgery needs robust and beneficiary control techniques for providing efficient and precise alternative for medical surgeries. Time delays over network severely impact the stability in the surgery and eventually affect the performance of expert doctors. Also we analyze the problems encountered with the M2M connectivity of network devices.

show abstract

A Comparative Study of Classification Techniques for Managing IoT Devices of Common Specifications

Cited by 15 publications

References 12 publications

Unsupervised Learning for Trustworthy IoT

Unsupervised Learning for Trustworthy IoT

A Conceptual Architecture in Decentralizing Computing, Storage, and Networking Aspect of IoT Infrastructure

Device Classification for Machine to Machine Communication in Internet of Things for Tele-Robotic Surgery: A Review

Contact Info

Product

Resources

About