An Inclusive Analysis on Deep Learning Hinged Malware Detection Techniques

An Inclusive Analysis on Deep Learning Hinged Malware Detection Techniques

연구 분야: Safety

학회: International Conference on Artificial Intelligence and Speech Technology

Malwares are touching the sky in connection with their numbers and this accretion is contemporaneous with the amelioration in usage of Android smart phones. Nearly everything is a blessing and curse at the same time. This applies to innumerable applications available in play stores too. Malware developers earmark this platform to consummate their thrust. These malicious programs can torment the devices in disparate ways such as absconding with the privileged information, superintending the device, wreaking havoc on the battery and maneuvering the data stockpiled in the device. Malware has become proficient in coping with various customary detection techniques. Static detection methods are powerless to seize the malware that utilizes dynamic features such as network traffic features whereas dynamic detection procedures discover it strenuous to detect malware that lodges static features like permissions, images and intents etc. Therefore, the exigency surfaces to maneuver some advanced methodologies in this direction. Deep learning coupled with numerous techniques fulfill this objective. This paper analyzes numerous deep learning techniques which are worthwhile for the detection of these stubborn malwares. The analysis throws light on various deep learning models also.