Kadam, Sonali Ramesh (2024) Exploring comprehensive analytical approaches for blockchain network data fraud detection. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (111kB) | Preview |
Abstract
This research focuses on logical analysis of Blockchain network transactions by using the concepts of Machine Learning incorporated to improve the accuracy & security aspects in financial systems. In meeting these research questions, this study adopts sturdy models like linear regression, random forest, and anomaly detection to deal with problems arising from transactions and, more specifically, fraud in the cryptocurrency space.
Chapter 1 introduction, this chapter presents the importance of focusing on blockchain technology in the context of the financial industry, explaining the possible impact on the transactional phase. It presents the rescission objectives; These are the research questions: While building the predictive models, and investigating the security in the blockchain networks.
Chapter 2 moves around related works where various literatures on blockchain technology and more specifically, the application of machine learning in blockchain, shall be reviewed. This is done by presenting different methodologies in predictive modelling and presenting a rationale in the use of anomaly detection when screening for fraud, thus laying down a conceptual background of the research work.
Chapter 3, methodology, describes the methods used in developing the research, data acquisition, data preparation, and selection of the machine learning algorithm. This is a basic approach to the organisational goals of linear regression, random forests, and anomaly detection algorithms and provides the foundation for analysis.
Chapter 4 design specifications this chapter outlines the process followed in the analysis of the transaction datasets through various data transformations on the datasets. Meaning it comprises data preprocessing and cleaning where there is the management of missing values, feature engineering and exploratory data analysis that is a way of preparing data for the subsequent modelling by ensuring high data quality.
Chapter 5 implementation, entails advancement in explaining the process of applying the adopted machine learning techniques to the blockchain data. Also described are how the different libraries were applied, the characteristics of the datasets employed, and the assessment of how well models perform, all of which show how well the strategies applied performed in this context.
Chapter 6 evaluation evaluates the efficiency of the models, which predict the commitment of the transactions, and the goal of processing the transactions within the Blockchain related network. It uses actual and predicted values to draw its conclusions, reviews the scalability issue and the necessity of anomaly detection for improving security of financial transactions.
Chapter 7 conclusion and future work provides a brief on the main conclusions on the given subtopic and its significance in the future of blockchain analytics. It relates to the research objectives and describes the scope of the study’s limitations while also providing a glimpse into future work, which requires extended and enriched data sets, greater processing immediacy
Actions (login required)
![]() |
View Item |