International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 9 | Sep-2024
p-ISSN: 2395-0072
www.irjet.net
Efficient method for Android Malware Family Detection and Classification using Sequential Network with ECLAT Ayush Wardhan sahu1, Prof. Satendra Sonare2 1Reseacrh Scholar, Department of CSE, Gyan Ganga Institute of Technology and sciences, Jabalpur, M.P. 2Professor, Departmet of CSE, Gyan Ganga Institute of Technology and Sciences, Jabalpur, M.P.
---------------------------------------------------------------------***-------------------------------------------------------------------same malware. However, it failed to detect previously Abstract – Android's popularity makes it a lucrative target
unseen malware. Although detection approaches based on behavior, heuristics, and model checking are effective in detecting some parts of unknown malware, they cannot show the same performance in detecting complex malware variants that use obfuscation and packaging techniques. A deep learning approach is starting to be used as a new paradigm to overcome the shortcomings of existing malware detection and classification approaches. Deep learning has been widely used in various fields including image processing, computer vision, human activity recognition [3], driving safety [4], facial emotion recognition [5], and natural language processing. However, it is little used in the field of cyber security, especially in malware detection. Deep learning is a subset of artificial intelligence that operates on artificial neural networks (ANNs). Deep learning uses multiple hidden layers and learns from examples. Recently, several deep learning architectures such as deep neural networks (DNNs), deep belief networks (DBNs), recurrent neural networks (RNNs), and convolutional neural networks (CNNs) have been used to improve model performance. Deep learning brings many advantages over the traditional learning scheme:
for cybercriminals. Malware with different behavior patterns that specifically target user routines is constantly entering the market. For this reason, knowing how to identify different forms of malware is essential to anti-malware protection. Android malware has become a serious threat to our daily lives, so it is urgently necessary to effectively mitigate or defend against it. Recently, many Android malware analysis approaches and tools have been proposed to protect legitimate users from the threat. However, most approaches focus on malware detection, while only a few consider malware classification or malware characterization. This paper proposes the use of ECLAT-based machine learning and deep learning methods to classify malware families and categories based on many different datasets in order to evaluate and select appropriate methods for each dataset.
Keywords: Android malware, Malware Features, Deep Learning, ECLAT, Sequential Neural Network, Accuracy.
1. INTRODUCTION To protect computer systems, we need to detect malware as soon as it infects systems. Malware detection is the process of analyzing a suspicious file and identifying whether it is malware or benign. Malware classification is a step further. After a file is identified as malware, by entering a category or group of malware known as a malware classification. Malware detection requires 3-step operations:
1. The DL model can automatically generate high-level elements from existing elements. 2. DL reduces the need for feature engineering. 3. DL can handle unstructured data efficiently. 4. DL can handle very large datasets.
1. Malware files are analyzed using appropriate tools. 2. Static and dynamic elements are extracted from the analyzed files.
5. DL reduces the element space. 6. DL can effectively perform unsupervised, semi-supervised and supervised learning.
3. Functions are grouped in certain ways to separate malicious software from benign software. Various sciences and techniques including data science, machine learning, and heuristics, as well as technologies such as cloud computing, big data, and blockchain, are used in these processes to increase detection rates. There are different approaches to detect malware using the above techniques and technologies. These approaches are mainly signature, behavior, model checking and heuristic detection [1], [2]. The names of these approaches vary according to the techniques and technologies used. The signature-based approach is effective for known and similar versions of the
© 2024, IRJET
|
Impact Factor value: 8.315
7. DL reduces costs and increases accuracy. 1.2 Types of Malware Malware, Short For Malicious Software, Is Software Designed by Cyber Attackers to Access or Harm a Computer or Network without the Victim’s Knowledge. Malware Is Defined As Any Software Designed To Cause Direct Harm. Despite A 39% Decrease In Global Malware Volume In 2020, Malware Attacks Continue. At The Same Time, Some Types
|
ISO 9001:2008 Certified Journal
|
Page 116