Make ML outcomes roar with graph neural networks on TigerGraph and AWS (PRT097)
summary
TigerGraph claims its distributed graph DB hits a 36 terabyte benchmark (in‑memory) and runs graphs with up to 10 billion vertices and 50 billion edges, exposes a Python ML Workbench with 50+ graph algorithms, and plugs into Amazon SageMaker to train GNNs. Customers report ~20%+ AUC improvement over existing models and some financial firms attribute over $100M annual benefit from improved fraud / AML detection; examples include supply‑chain cuts from weeks to 45 minutes.
AI-written summary · about
chapters (5)
- 0:00Start
- 1:04Real life
- 3:39Throwing half away
- 4:14Graph databases
- 7:07Graph features boost machine learning
related talks
28mA Quantum Leap In Fraud Detection Accuracy Using Graph-based Features
29mTigerGraph, Connecting the World
23mGraph ML – The Next Level Of Machine Learning; Learn How To Use It And When Not To
49mModern Graph Databases
26mAdvancing Financial Fraud Detection with Graph Dbs
27mData @scale Seattle - Yucheng Low
31mGraph-Based Machine Learning with Python
56mGraph Analytic Systems