New AI framework FraudShield AI enhances financial fraud detection

Researchers have developed FraudShield AI, a novel framework designed to combat sophisticated financial fraud like money laundering. This system combines Long Short-Term Memory (LSTM) networks with graph neural networks to analyze both temporal transaction sequences and relational network structures. By incorporating features such as PageRank centrality and flow ratios, FraudShield AI aims to detect subtle, network-level fraudulent activities that traditional methods might miss. Experiments on the PaySim dataset demonstrated that this hybrid approach significantly outperforms baseline models like Logistic Regression and XGBoost in identifying micro-transaction fraud. AI

IMPACT This framework could improve the accuracy and resilience of financial fraud detection systems against evolving adversarial tactics.

RANK_REASON The cluster contains an academic paper detailing a new AI framework for a specific application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Focal loss
• FraudShield AI
• Graph Topological Features
• logistic regression model
• long short-term memory
• PaySim
• XGBoost

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →