Adversarial Machine Learning in e-Health: Attacking a Smart Prescription System

Gaglio, Salvatore; Giammanco, A.; Re, Giuseppe Lo; Morana, Marco

doi:10.1007/978-3-031-08421-8_34

Cited by 3 publications

(1 citation statement)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A close look at these applications reveals current state of cloud computing in e-health platforms and compares advantages of these systems. cloud service in an e-health system A broker-based access control scheme for selective EHR sharing among different healthcare providers in the cloud was proposed in the work [7]. In addition, they did not consider how to implement their sharing idea on resource-constrained devices like smartphones, instead focusing only on a computer simulation using VMs.…”

Section: Related Workmentioning

confidence: 99%

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

Awasthi

Suchithra

Chakravarty

et al. 2023

Preprint

View full text Add to dashboard Cite

Numerous aspects of healthcare have been altered by cloud-based computing. Scalability of required service as well as ability to upscale or downsize data storage, as well as the collaboration between AI and machine learning, are main benefits of cloud computing in healthcare. Current paper looked at a number of different research studies to find out how intelligent techniques can be used in health systems. The main focus was on security and privacy concerns with the current technologies. This study proposes a novel method for cloud service device-to-device communication using feature selection and classification for data analysis in an e-health system. Through a comprehensive requirement analysis as well as user study, the purpose of this research is to investigate viability of incorporating cloud as well as distributed computing into e-healthcare. After that, the smart healthcare system and conventional database-centric healthcare methods will be compared, and a prototype system will be created as well as put into use based on results. Convolutional adversarial neural networks with transfer perceptron are used to analyze the cloud-based e-health data that has been collected. Proposed technique attained training accuracy 98%, validation accuracy 93%, PSNR 66%, MSE 68%, precision 72%, QoS 63%, Latency 58%.

show abstract

Section: Related Workmentioning

confidence: 99%

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

Awasthi

Suchithra

Chakravarty

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

Awasthi,

Suchithra,

Chakravarty

et al. 2023

Soft Comput

View full text Add to dashboard Cite

AdverSPAM: Adversarial SPam Account Manipulation in Online Social Networks

Concone,

Gaglio,

Giammanco

et al. 2024

ACM Trans. Priv. Secur.

View full text Add to dashboard Cite

In recent years, the widespread adoption of Machine Learning (ML) at the core of complex IT systems has driven researchers to investigate the security and reliability of ML techniques. A very specific kind of threats concerns the adversary mechanisms through which an attacker could induce a classification algorithm to provide the desired output. Such strategies, known as Adversarial Machine Learning (AML), have a twofold purpose: to calculate a perturbation to be applied to the classifier’s input such that the outcome is subverted, while maintaining the underlying intent of the original data. Although any manipulation that accomplishes these goals is theoretically acceptable, in real scenarios perturbations must correspond to a set of permissible manipulations of the input, which is rarely considered in the literature. In this paper, we present AdverSPAM , an AML technique designed to fool the spam account detection system of an Online Social Network (OSN). The proposed black-box evasion attack is formulated as an optimization problem that computes the adversarial sample while maintaining two important properties of the feature space, namely statistical correlation and semantic dependency . Although being demonstrated in an OSN security scenario, such an approach might be applied in other context where the aim is to perturb data described by mutually related features. Experiments conducted on a public dataset show the effectiveness of AdverSPAM compared to five state-of-the-art competitors, even in the presence of adversarial defense mechanisms.

show abstract

Adversarial Machine Learning in e-Health: Attacking a Smart Prescription System

Cited by 3 publications

References 23 publications

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

Machine learning-based D2D communication for a cloud-secure e-health system and data analysis by feature selection with classification

AdverSPAM: Adversarial SPam Account Manipulation in Online Social Networks

Contact Info

Product

Resources

About