Machine Learning with Bitcoin Heist Ransomware

Abstract

In recent years, there has been a significant rise in the popularity of cryptocurrency amongst investors worldwide. One cryptocurrency that has been the forerunner in this new digital age is Satoshi Nakamoto’s Bitcoin. As much as it has augmented in value in the past several years, many issues have emerged as new points of concern surrounding cryptocurrency ransomware orchestrated by scammers. As a result of the growing scandals, one notorious case that has made the most headlines is the Bitcoin Heist. We have found a sizable dataset that traces back to the Bitcoin Heist incident. With the help of data science and machine learning fundamentals, we will explain different methodologies to determine whether transactions are malicious or not based on a given Bitcoin address. In this paper, we will explain cryptocurrency and ransomware and further insights into the machine learning concepts behind this issue through various models such as Adaptive Boosting (AdaBoost), Gradient Boosting, K-Nearest Neighbor (KNN), and Random Forest.