Live Demo: View Project Details on Power BI
Following the recent significant technological advancement, fraudsters have become increasingly easier to committ fraud schemes with tremendous impacts on businesses. It is therefore crucial for businesses to gear themselves up to face such challenges ahead by using various means in data analytics. In this project on fraud detection analysis, Power BI has been used to visualize the basic features of the PaySim synthetic dataset on the routine operation of financial transactions and their associated, if any, malicious behaviour.
This project begins by showing various forms of visualizations relating to the basic features of the dataset such as the count of different transaction type, transaction amount by types, transaction amount against time steps, the diffferences in account balance between the original and destination accounts in the case of the original and destination accounts respectively. It follows by showing certain descriptive statistics about the fraud and non=fradulent cases respectively such as the fraud count by transaction types, non-fraud count against the transaction amount, account balance differnces in the fraudulent cases and non=fraudulent cases, etc.
Lastly, the project ends up with an automated, chatbot-like Q&A session.