The interconnectivity, accessibility and open nature of IT industry has proved to be a boonfor both developers and users.
But it comes with some threats as well. The mostsignificant one is the spread of malwares. Malware referred to as Malicioussoftware in any software application that can infiltrate into a system andaccess or damage resources without the owner’s consent. Malware is a genericterm that may be viruses, worms, Trojan horses, spyware etc.· Adware -These are malwares which automatically show the advertisement to the user.· Virus – It is the software which can harm yourcomputer by generating its copy automatically. These can be sent throughelectronic mails, files etc.
· Worm –They can be sent with the help of networks. They have tendency to self-replicateitself and disseminate independently. On the other hand, viruses spread whenthe user take part in this activity.
· Backdoors– These are the software’s which bypass the login credentials without detectedby the owner. One or more software’s can be installed into system for futureuse. The potential harm that may result from themalware requires the anti-malware authors to stay a step ahead of the malwareauthors. This paper describes the use of LeNet like convolution neural networkfor malware detection. Malware detection is a technique that is used todistinguish between a malicious application from a benign one. Not only this,as there are lots of categories of malwares, malware classification is alsoimportant.
I. ChalLEnges in malwre detectionIn present scenario we detect the malwares by signaturebased methods and this process has been used by antivirus vendors from lastmany years. Malware signature is a kind of algorithm which helps us to identifythe type of the malware. When we identify the malware then it is not easy toidentify its family as hackers use the polymorphic engine and metamorphicengine to stay step ahead form the anti-virus programmers. Lack of open sourcedataset for malware poses a great challenge since success of a machine learningalgorithm largely depends on the quantity and quality of the dataset used.
Newmalwares get inflected into the system with every tick of the clock. Malwaredetection suffers with the problem akin to the problem in virus detection in biologicalsystems. The files look different but actually belong to the same family.
Themalware authors use polymorphism by virtue of which the same binary file aremodified such that they look completely different. This makes use oftraditional techniques inefficient. Another challenge is the large number