Quadratic multivariate linear regressive distributed proximity feature engineering for cybercrime detection in digital fund transactions with big data

Digital fund transactions involve the electronic transfer of funds between parties through digital channels such as online banking platforms, mobile applications, and electronic payment systems. However, the rapid advancement of digital transactions has also directed cybercriminals to exploit vulnerabilities, engaging in money laundering and other illegal activities, resulting in substantial financial losses. The improve accuracy of cybercriminal detection by lesser time consumption, a novel technique called quadratic multivariate linear regressive distributed proximity feature engineering (QMLRDPFE) is developed. The proposed QMLRDPFE technique comprises two primary steps namely data preprocessing and feature engineering. Analyzed results prove that the QMLRDPFE technique outperforms existing methods in attaining superior accuracy and precision. Furthermore, QMLRDPFE method shows effective in reducing time utilization and space complexity for fraudulent transaction detection compared to existing approaches. Results to provide effective in reducing time utilization and space complexity for fraudulent transaction detection than the conventional methods.