XTRACTIS for fraud detection

Physicochemical Identification of Wines' Origin

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

Design an AI-based decision-making system that accurately identifies a wine from its physico-chemical analysis, in order to detect fraudulent bottles of wine, in a rational and explainable way, and help preserve the wine’s protected designation of origin.

Goals & Benefits

  1. Identify the specific physicochemical parameters significantly characterizing each wine and enhance expert knowledge by helping wine-specializing chemists understand the cause-and-effect relationships between these parameters and the wine origin.
  2. Help anti-fraud services detect counterfeit wines and understand the underlying strategy of fraudsters in the way they counterfeit wine.
  3. Avoid false alarms thanks to transparent and accurate automatized wine composition analysis.

XTRACTIS-induced Decision System

Intelligible Model, Explainable Decisions
  • The top-model is a decision system composed of 7 rules without chaining, 3 of which are gradual.
  • Each rule uses from 2 to 3 predictors among the 8 variables that XTRACTIS automatically identified as significant in the decision process (out of the 13 Potential Predictors).
  • Only a few rules are triggered at a time to compute the decision

It has a perfect Real Performance (on unknown data).

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

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

Detailed results and explanations in full document

Use Case 2025/09 (v1.0)

Powered by XTRACTIS® REVEAL v14.0.56860 (2025/09)

Download Document

CONTENTS

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