Abstract

Artificial intelligence (AI) evaluates networking data to detect and prevent data leakage, while enhancing security in financial cloud networks through advanced optimisation. Consumption of resources, scalability issues, concerns about privacy, false-positives and adapting to emerging threats reduce the performance of the data leakage detection (DLD) system. In this investigation, research proposed a novel model of gravitational search inspired flexible Adaboost (GS-FAB) for developing a DLD framework to improve financial cloud networking security by proactively identifying and mitigating risks. The financial cloud network security dataset was collected in this research. After the data collection, we applied a robust scaler for data preprocessing. After data preprocessing, the data labelling process is implemented and that could be divided into non-leakage and leakage data. Following the leakage and non-leakage data process, the research separated the leakage and non-leakage data into training and testing. The training process includes the GS-FAB technique. The GS-FAB method and testing process encompass the performance analysis. Finally, research employs some metrics for existing and proposed methods. Our proposed GS-FAB technique results include precision (95%), f1-score (90.5%), recall (91.4%), RMSE (49.4%), SNR (64%) and accuracy (98.1%). In comparison to both proposed and existing approaches, our suggested GS-FAB technique obtained better results. This investigation not only enhances the accuracy but also promises a better effective technique for improving financial cloud networking security by providing robust protections from unwanted data accessibility.