XTRACTIS for Security & Smart Cities

Prediction of Robbery Crimes in American Cities

Benchmark vs. Random Forest, Boosted Tree & Neural Network

Design an AI-based decision system that accurately predicts the number of robberies committed in a year in a city, given the hyper-complexity of the phenomenon (highly nonlinear behavior) in order to identify sources of crime and warn about the criminality level in a rational and explainable way.
Goals & benefits

Allow city administrators and police departments to understand the causal relationships between socio-economic factors or security policy and future criminal activities.

Find the truly influential parameters for assessing criminality to define an effective city policy.

Carry out dedicated police actions to lower criminality, in accordance with population or macro-economic changes.

  • The top-model is a decision system composed of 18 gradual rules without chaining.
  • Each rule uses from 2 to 4 predictors among the 14 variables that XTRACTIS automatically identified as significant (out of the 102 features characterizing robbery crime records).
  • Only a few rules are triggered at a time to compute the decision.

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

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

UC23 scores graph

RFo=Random Forests
BT=Boosted Trees
NN=Neural Networks

Detailed results and explanations in full document

Use Case 2024/06 (v1.1)

Powered by XTRACTIS® REVEAL 13.0.47764 (2023/11)


  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