Since the emergence of blockchain technology in the form of Bitcoin, its development has progressed rapidly and attracted the attention of various researchers in academia and industry. Blockchain technology is becoming an increasingly secure and effective way to share information in multiple sectors, including finance, healthcare, supply chain management (SCM), and the Internet of Things (IoT). The decentralized and anonymous characteristics of transactions on a cryptocurrency blockchain have drawn many participants, resulting in substantial daily monetary exchanges. This necessitates the examination of the blockchain to uncover information about the characteristics of individuals involved in transactions. This research aims to develop a machine learning-based model for classifying blockchain transactions into risky or non-risky ones. An ensemble machine learning model using the voting technique with an ensemble feature selection method is developed with the classifiers Support Vector Machine (SVM), XGBoost, and AdaBoost. The proposed model is evaluated using Ecliptic++ blockchain transactional data. The empirical analysis done in this research shows that the ensemble model with hard voting has an accuracy of 90.13%, and soft voting has an accuracy of 88.48%.