Student Name

Ramsha ArifFollow

Degree

Master of Science in Data Science

Department

Department of Computer Science

Faculty/ School

School of Mathematics and Computer Science (SMCS)

Date of Submission

Spring 2022

Supervisor

Dr. Imran Khan, Assistant Professor, Department of Computer Science

Abstract

Despite the fact that the proliferation of cryptocurrencies such as Bitcoin, that allow pseudo-anonymous transactions have eased the path of cybercriminals by digitally transferring money without revealing their identity and using it for illegal purposes. Nowadays, ransomware attackers block access to sensitive user data and extort ransom in Bitcoin. Although the transaction records are publicly available, current data analytic approaches that depend on machine learning techniques to detect ransomware addresses have shown results below par. Using the existing ransomware dataset, this research applied an ensemble machine learning technique to detect new malign addresses, given previous partial transaction records, the ensemble technique provides significantly improved outcome with 86% of recall value. Furthermore, a hybrid blockchain-based federated learning framework is proposed to incorporate the aforesaid technique on cryptocurrency wallets. While governments and financial bodies are being cautious and have not accepted cryptocurrencies as a legal tender, this research will help to plug in the proposed framework in any type of cryptocurrency.

Project Type

MSCS Survey Report

Document Type

Thesis

The full text of this document is only accessible to authorized users.

Share

COinS