GRAPH ALGORITHMS · C++

Fraud Detection with PPR

Personalized PageRank for scalable ranking-based fraud detection in transaction networks.

Role Algorithms / Systems

Year 2026

Language C++17 + Python

Evaluation Convergence · Precision@K

View source on GitHub ↗ Back to work

Fraud Detection with PPR convergence analysis

Graph-based fraud detection

A performance-oriented C++17 implementation of Personalized PageRank using power iteration, deterministic seed selection, convergence monitoring, Precision@K evaluation, and structured experiments across datasets and alpha values.

Graph Input Seed Selection Power Iteration Convergence Ranking Precision@K Parameter Sweep Analysis

Parameter sensitivity