Malware Detection with Deep Neural Network Using Process Behavior

Tobiyama, Shun; Yamaguchi, Yukiko; Shimada, Hiroyuki; Ikuse, Tomonori; Yagi, T.

doi:10.1109/compsac.2016.151

Cited by 242 publications

(140 citation statements)

References 4 publications

Supporting

Mentioning

135

Contrasting

Unclassified

Order By: Relevance

“…To align the sequences a warping matrix has been constructed and search for the optimal warping path (red/solid squares) as shown on the right figure with the Sakoe-Chiba Band with width R is used to constrain the warping path. This supports the idea that Dynamic time warping is an algorithm for measuring the similarity of two-time series (Tobiyama et al, 2016). …”

Section: Dynamic Time Warping (Dtw)supporting

confidence: 82%

“…The only restriction placed on the data sequences is that they should be sampled at equidistant points in time (this problem can be resolved by re-sampling) (Senin, 2008). The below figure is taken from (Tobiyama et al, 2016) explains more in the concept of DTW. Figure 6 on the left shows the two-time series which are similar but out of phase and has produced a large Euclidean distance.…”

Section: Dynamic Time Warping (Dtw)mentioning

confidence: 99%

See 1 more Smart Citation

Survey on Representation Techniques for Malware Detection System

Mohamed

Ithnin

2017

American Journal of Applied Sciences

View full text Add to dashboard Cite

Malicious programs are malignant software's designed by hackers or cyber offenders with a harmful intent to disrupt computer operation. In various researches, we found that the balance between designing an accurate architecture that can detect the malware and track several advanced techniques that malware creators apply to get variants of malware are always a difficult line. Hence the study of malware detection techniques has become more important and challenging within the security field. This review paper provides a detailed discussion and full reviews for various types of malware, malware detection techniques, various researches on them, malware analysis methods and different dynamic programmingbased tools that could be used to represent the malware sampled. We have provided a comprehensive bibliography in malware detection, its techniques and analysis methods for malware researchers.

show abstract

Section: Dynamic Time Warping (Dtw)supporting

confidence: 82%

Section: Dynamic Time Warping (Dtw)mentioning

confidence: 99%

Survey on Representation Techniques for Malware Detection System

Mohamed

Ithnin

2017

American Journal of Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Second, in the testing phase, obtained predictive model process the properties of unknown data and predict whether it is malware or benign. The popular machine learning techniques among the researchers for the detection of second generation malware are Naive Bayes (Sharma, Sahay, & Kumar, ) (Shabtai, Kanonov, Elovici, Glezer, & Weiss, ), Decision Tree (Sharma & Sahay, ), Data Mining (Santos, Brezo, Ugarte‐Pedrero, & Bringas, ), Neural Networks (Dahl, Stokes, Deng, & Yu, ), Hidden Markov Models (Raghavan, ), Deep Learning (Tobiyama, Yamaguchi, Shimada, Ikuse, & Yagi, ), etc.…”

Section: Applications Of Machine Learning In Cybersecuritymentioning

confidence: 99%

Machine learning in cybersecurity: A review

Handa

Sharma

Shukla

2019

WIREs Data Min & Knowl

103

View full text Add to dashboard Cite

Machine learning technology has become mainstream in a large number of domains, and cybersecurity applications of machine learning techniques are plenty. Examples include malware analysis, especially for zero‐day malware detection, threat analysis, anomaly based intrusion detection of prevalent attacks on critical infrastructures, and many others. Due to the ineffectiveness of signature‐based methods in detecting zero day attacks or even slight variants of known attacks, machine learning‐based detection is being used by researchers in many cybersecurity products. In this review, we discuss several areas of cybersecurity where machine learning is used as a tool. We also provide a few glimpses of adversarial attacks on machine learning algorithms to manipulate training and test data of classifiers, to render such tools ineffective. This article is categorized under: Application Areas > Science and Technology Technologies > Machine Learning Technologies > Classification Application Areas > Data Mining Software Tools

show abstract

“…For example in [9], time series data from smart grids were analyzed for electricity theft detection, where in such use cases perturbed data can help thieves to avoid being detected. Other crucial decision making systems such as malware detection in smart-phones, leverage temporal data in order to classify if an Android application is malicious or not [5]. Using adversarial attacks, a hacker might generate synthetic data from his/her application allowing it to bypass the security systems and get it installed on the end user's smart-phone.…”

Section: A Deep Learning For Time Series Classificationmentioning

confidence: 99%

“…Time Series Classification (TSC) problems are encountered in various real world data mining tasks ranging from health care [1]- [3] and security [4], [5] to food safety [6], [7] and power consumption monitoring [8], [9]. As deep learning models have revolutionized many machine learning fields such as computer vision [10] and natural language processing [11], [12], researchers recently started to adopt these models for TSC tasks [13].…”

Section: Introductionmentioning

confidence: 99%

Adversarial Attacks on Deep Neural Networks for Time Series Classification

Fawaz

Forestier

Weber

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

Time Series Classification (TSC) problems are encountered in many real life data mining tasks ranging from medicine and security to human activity recognition and food safety. With the recent success of deep neural networks in various domains such as computer vision and natural language processing, researchers started adopting these techniques for solving time series data mining problems. However, to the best of our knowledge, no previous work has considered the vulnerability of deep learning models to adversarial time series examples, which could potentially make them unreliable in situations where the decision taken by the classifier is crucial such as in medicine and security. For computer vision problems, such attacks have been shown to be very easy to perform by altering the image and adding an imperceptible amount of noise to trick the network into wrongly classifying the input image. Following this line of work, we propose to leverage existing adversarial attack mechanisms to add a special noise to the input time series in order to decrease the network's confidence when classifying instances at test time.Our results reveal that current state-of-the-art deep learning time series classifiers are vulnerable to adversarial attacks which can have major consequences in multiple domains such as food safety and quality assurance.• The definition and formalization of adversarial attacks for

show abstract

Malware Detection with Deep Neural Network Using Process Behavior

Cited by 242 publications

References 4 publications

Survey on Representation Techniques for Malware Detection System

Survey on Representation Techniques for Malware Detection System

Machine learning in cybersecurity: A review

Adversarial Attacks on Deep Neural Networks for Time Series Classification

Contact Info

Product

Resources

About