Optimization of Anti‐Tampering Method for Financial Data in Edge Intelligent Systems Based on Support Vector Machine

ABSTRACT The proliferation of edge intelligent systems has introduced new challenges in securing financial data across distributed networks. This paper proposes an optimized anti‐tampering method for financial data based on SVM, specifically designed for implementation in edge computing environments. Our approach leverages the strengths of SVM in handling non‐linear, high‐dimensional data, while addressing the unique security and privacy concerns of edge‐based financial systems. We introduce a novel feature weighting mechanism that enhances the SVM's ability to detect subtle data tampering attempts in distributed financial datasets, and improve the application of SVM to better handle non‐stationary, asymmetric, and non‐linear time series financial data in edge computing settings. A comprehensive evaluation using real‐world financial data in simulated edge computing scenarios demonstrates significant improvements in yield indicators, maximum drawdown rate, and volatility compared to traditional approaches. Our method not only effectively detects data tampering but also improves the accuracy of financial data prediction in decentralized environments. This research contributes to the development of more secure, efficient, and privacy‐preserving financial systems in the increasingly decentralized global economy, paving the way for robust security solutions in edge‐based financial infrastructures.