Deep Learning Applications to Online Payment Fraud Detection
summary
Describes applying deep neural networks to online payments fraud detection at scale, addressing challenges like high dimensionality, sparse and high-cardinality features, and temporal transaction data. Discusses specific deep learning methods used for different problem formulations, empirical comparisons of their results, and high-level runtime performance and deployment considerations for training and inference (as applied at PayPal).
AI-written summary · about
chapters (6)
- 0:00Start
- 3:24PayPal Ecosystem (1)
- 5:40Problem Formulation - Fraud Detection
- 12:21Addressing Class Imbalance
- 24:44Reducing Good User Declines (2) Transfer Learning using Generative Modeling Context
- 27:33Deep Learning Applications to Fraud Detection Conclusions
related talks
18mReal-time Fraud Detection using Deep Learning
52mModern Fraud Prevention Using Deep Learning
45mAnti-Fraud in Reverse and the Use of ML Methods in It
26mAI-Powered Fraud Detection in Cloud-Native Digital Payments
35mAI Powered Fraud Detection for Payments & Transactions
37mThe Role of Machine Learning in Today’s Payment Ecosystems
16mLeveraging ML to Create a Reliable Anti-Fraud System
39mThe Science Behind Delightful User Experience