Abstract:
"
Ransomware is one of the most spreading and dangerous kinds of computer malware
these days. Attackers are using faster algorithms that have the capability to encrypt
files in few seconds. A cryptographic based ransomware attack cannot be decrypted
without the encryption key. Attackers are using social engineering techniques to
deliver and execute ransomware files on computers. In 2016, 200,000 computers
worldwide were got attacked by ransomware called WnnaCry. Large organizations
and many numbers of hospitals were lost their data at that time. Operating systems
default firewalls and anti-malware tools are not capable of identifying these attacks in
real-time.
In this research, the author has proposed an open-source solution to detect ransomware
attacks using machine learning. An Artificial Neural Network was built to identify
attack by analysing system activities. The system provides a faster detection method
to identify attacks. GigaRanD system is working with all the Windows 10 64-bit
operating system versions. The system is faster than currently available commercial
ransomware detection software also. The GigaRanD system is benchmarked with
worlds top rated anti-malware tools such as Kaspersky and ESET. The tool was 4 times
faster than the existing systems available. Since the system is an open-source project,
future researchers and developers can contribute to the system to improve accuracy
and performance"
