XTRACTIS FOR fraud detection

Detection of Fraudulent Credit Card Transactions

Benchmark vs. Logistic Regression, Random Forests, Boosted Trees & Neural Networks

Design, from only some transactions characteristics, an AI-based decision system that efficiently diagnoses credit card transactions, in order to rationally and instantly detect the fraudulent ones and to eventually adapt protection measures.
Goals & benefits

Identify the parameters involved in the fraudulent transactions and enhance knowledge by helping banking specialists understand the causal relationships between these parameters, their combination, and the occurrence of fraud (i.e., understand the scammers’ strategies).

Help the banking sector to make transparent decisions through automatic, explainable detection, while improving the consumer experience.

Use a detection model with fewer transaction characteristics to speed up protection process.

  • The top-model is a decision system composed of 27 gradual rules without chaining.
  • Each rule uses from 2 to 5 predictors among the 9 variables that XTRACTIS identified as significant (out of the 29 ones characterizing transactions).
  • Only a few rules are triggered at a time to compute the decision.

It has a very good Real Performance (on unknown data).

It computes real-time predictions up to 70,000 decisions/second, offline or online (API).

UC20 scores graph
LoR=Logistic Regression
RFo=Random Forests
BT=Boosted Trees
NN=Neural Networks

Detailed results and explanations in full document

Use Case 2024/02 (v2.1)

Powered by XTRACTIS® REVEAL 12.2.45294 (2023/06)


  1. Problem Definition
  2. XTRACTIS-induced Decision System
  3. XTRACTIS Process
  4. Top-Model Induction
  5. Explained Predictions for 2 unkown cases
  6. Top-Models Benchmark
  7. Appendices