Ransomware classification by machine learning and dimensionality reduction
Ransomware is a type of malware that aims to take control of systems that host or encrypt data - until a ransom is paid. This threat, which has been seen as a new type of terrorism, is a difficult task given the rapid spread and changes developers apply to encryption techniques. Given this, Machine Learning (ML) classifier algorithms have been reported as promising tools for classifying ransomware. This work explores 7 ML techniques in order to make 5 types of approaches, along with 2 dimensionality reduction techniques. The Gaussian Process presented the best performance, as it proved to be effective in four approaches.
Ransomware classification by machine learning and dimensionality reduction
-
DOI: 10.22533/at.ed.3172252201116
-
Palavras-chave: ransomware, machine learning, dimensionality reduction
-
Keywords: ransomware, machine learning, dimensionality reduction
-
Abstract:
Ransomware is a type of malware that aims to take control of systems that host or encrypt data - until a ransom is paid. This threat, which has been seen as a new type of terrorism, is a difficult task given the rapid spread and changes developers apply to encryption techniques. Given this, Machine Learning (ML) classifier algorithms have been reported as promising tools for classifying ransomware. This work explores 7 ML techniques in order to make 5 types of approaches, along with 2 dimensionality reduction techniques. The Gaussian Process presented the best performance, as it proved to be effective in four approaches.
- George Tassiano Melo Pereira
- Claudomiro de Souza de Sales Júnior
