XTRACTIS for Civil Engineering

Prediction of the Compressive Strength of Concrete

Benchmark vs. Random Forests, Boosted Trees & Neural Networks

Design an AI system that accurately and rationally models the compressive strength of high-performance concrete from its age, formulation, and some of its manufacturing characteristics, given the hyper-complexity of the phenomenon.
Goals & benefits

Allow domain experts and civil engineers to understand the causal relationships between concrete parameters and its compressive strength.

Find the truly influential parameters to anticipate the compressive strength of the concrete and thus find better formulations or optimize its production.

Create new custom-designed concretes for specific uses.

  • The top-model is a decision system composed of 86 gradual rules without chaining.
  •  Each rule uses from 3 to 8 predictors among the 8 potential predictors that XTRACTIS identified as all significant.
  • The model is quite intelligible despite the large number of rules, given the high complexity of this regression modeling problem.
  • 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).

UC15 scores graph

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

Detailed results and explanations in full document

Use Case 2024/03 (v2.0)

Powered by XTRACTIS® REVEAL v12.2.44127 (2023/01)


  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