A Gaussian mixture model for dynamic detection of abnormal behavior in smartphone applications

Attar, Ali El; Khatoun, Rida; Lemercier, Marc

doi:10.1109/giis.2014.6934278

Cited by 10 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other studies using real devices for training and testing of models, based their performance on apps running isolated for 10 minutes [11], devices in an idle state [19] or unknown circumstances [13,28]. One study that used real devices under real-life circumstances for the assessment of their detection method is Reference [37].…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Dynamic Detection of Mobile Malware Using Smartphone Data and Machine Learning

2022

View full text Add to dashboard Cite

Mobile malware are malicious programs that target mobile devices. They are an increasing problem, as seen in the rise of detected mobile malware samples per year. The number of active smartphone users is expected to grow, stressing the importance of research on the detection of mobile malware. Detection methods for mobile malware exist but are still limited. In this paper, we propose dynamic malware-detection methods that use device information such as the CPU usage, battery usage, and memory usage for the detection of 10 subtypes of Mobile Trojans on the Android Operating System (OS). We use a real-life sensor dataset containing device and malware data from 47 users for a year (2016) to create multiple mobile malware detection methods. We examine which features, i.e. aspects, of a device, are most important to monitor to detect (subtypes of) Mobile Trojans. The focus of this paper is on dynamic hardware features. Using these dynamic features we apply the following machine learning classifiers: Random Forest, K-Nearest Neighbour, and AdaBoost.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Other research assessed have also used machine learning for the dynamic detection of mobile malware, but this literature is unclear about the feature collection method (Ham and Choi [28]), include few malware samples (Attar [13] and Dixon [23]), or experiment on an idle phone (Caviglione [19]).…”

Section: Related Workmentioning

confidence: 99%

Dynamic Detection of Mobile Malware Using Smartphone Data and Machine Learning

2022

View full text Add to dashboard Cite

show abstract

“…The procedure is described in Fig. 5 The values of the combination scenario can be calculated by (14).…”

Section: Nonlinear Two-cluster Gaussian Mixture Scenario Modelmentioning

confidence: 99%

“…However, they do not work for fitting the irregular wind speed probability distribution. The Gaussian mixture model (GMM) [12][13][14] can be used to fit any arbitrarily shaped probability distribution effectively by a finite number of Gaussian distribution functions, which can be applied to fit irregular wind speed probability distribution. On the other hand, the expectation maximization (EM) algorithm [15][16][17][18][19][20][21] is often used to construct the GMM, which was first proposed by Dempster, Laird, and Rubin (DLR) in 1977.…”

Section: Introductionmentioning

confidence: 99%

A nonlinear two‐cluster Gaussian mixture scenario model for wind power

Chen

Wang

2016

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

To overcome the low precision and poor flexibility of the wind power scenario model, this paper proposes a nonlinear twocluster Gaussian mixture scenario model for wind power based on the expectation maximization (EM) algorithm according to Wasserstein optimal scenario theory. First, the EM algorithm is used to classify the wind speed data and establish the Gaussian mixture model (GMM). Second, the Wasserstein distance scenarios of two Gaussian distributions with different parameters are calculated based on wind speed data and nonlinear wind turbine power curve, respectively. Finally, a cross combination of the two scenarios is used to obtain the nonlinear two-cluster mixture scenario model with a mixture Gaussian probability parameters. The obtained results show that the nonlinear two-cluster Gaussian mixture scenario model is applicable not only to the regular wind speed probability distribution but also to that with irregular and two-peak characteristics. Moreover, the generated energy deviation can be controlled within 2.5%.

show abstract