Reproducible research comparing GNN (GraphSAGE, GCN, GAT) vs ML baselines (XGBoost, RF) on Elliptic++ Bitcoin fraud detection. Features ablation experiments revealing when tabular models outperform graph neural networks.

A pytorch implementation of GCN on EllipticDataSet, and this repo provides a GCN-related layer to promote the original GCN layer.

Graph-Tabular Fusion for Bitcoin Fraud Detection - Demonstrating when Node2Vec embeddings don't improve XGBoost. Scientifically rigorous negative result validating that tabular features encode graph structure.

🔗 Explore Graph-Tabular Fusion to enhance Bitcoin fraud detection, revealing that rich tabular features render graph embeddings redundant.

Reproducible multi-dimensional framework for evaluating XAI methods on blockchain fraud detection: fidelity, stability, BRAS (domain alignment), and LLM-agent validation across Elliptic and Ethereum datasets.

Machine learning models for Anti-Money Laundering (AML) and fraud detection in cryptocurrency networks using graph-based transaction data.

End-to-end ML pipeline to detect illicit Bitcoin transactions using the Elliptic dataset. Includes clean training/evaluation CLI, SMOTE and CI (Black/Ruff + tests).

Detección de fraude en transacciones de Bitcoin usando el dataset Elliptic. Análisis exploratorio temporal (49 snapshots, 165 features), reducción de dimensionalidad (PCA/UMAP) y comparación de 8 modelos: Regresión Logística, Random Forest, SVM, XGBoost, MLP, GCN y GAT. El mejor rendimiento se logró con XGBoost (F1-score ~0.94).

End-to-end AML detection pipeline on the Elliptic Bitcoin transaction graph, comparing tabular baselines (cuML/sklearn) and GNNs (PyTorch Geometric) with reproducible phases, metrics, and curated results.