Robust strategies, intelligible models,
immediately operational solutions

Augmented Fuzzy Artificial Intelligence

The xtractis® robot benefits from the most recent advances in Fuzzy Logic coupled with Machine Learning. Using its unique mathematical approach and proprietary algorithms, xtractis® analyzes your multidimensional databases and automatically discovers decision-support systems, able to predict the process under study in the most reliable way.

The use of gradual and intelligible “If … Then” rules allows the domain expert to better understand the behavior of his/her process.

Predictive Robust Models

Thanks to its robust learning strategies, xtractis® “does not learn off by heart”: it builds and selects the decision-making models that have the best generalization capacity. xtractis® models hence have a better predictive power: they are effective even in unknown situations. They enable you to model and gain more insights into Real-World complexity without any simplification.

XTRACTIS®, the Intelligent Modeler Robot

xtractis® is fully autonomous thanks to an infinite number of robust learning strategies: it uses all the available data without any a priori knowledge.

All its functionalities are fully automated: launch xtractis® immediately by simple mouse click, without any need for programming or special knowledge of mathematics .

All types of complex processes

xtractis® is able to model expert decision-strategies, positive or malicious human behaviors, technical or socioeconomic processes or natural phenomena…

It identifies steady relationships between structured datasets of various types: qualitative and quantitative data, hedonic evaluations, liking, expert opinions, sensory/instrumental data, signals, images, socioeconomic data, formulations, product characteristics, genome sequences…

All types of modeling

xtractis® helps you to predict the value of a numerical variable (regression), to identify to which set a new observation belongs (classification), to predict the risk of occurrence of a studied event (scoring), or to identify stable clusters in your data (clustering).

Weak signals
are important

Some criteria may appear irrelevant when taken individually. The xtractis® approach is holistic: it highlights the existing synergies between these criteria and other parameters, to build the predictive model.

guarantees reliability

The xtractis® models are robust: they reflect stable behavior laws and ensure reliable predictions for situations that were not encountered during training.

Quickly straight to the target

The xtractis® models are deterministic: they always output the same result for a given situation. Predictions are instantaneous and optimal solutions are discovered in a few dozens of seconds.

Get your free white papers

Essay 1709 - pdg

Augmented Fuzzy AI vs. Connectionism (or xtractis® vs. Deep Learning)

PG - White Paper - intro to fuzzy logic EN

Introduction to Fuzzy Logic

PG - White Paper - xtractis Approach EN

How Augmented Fuzzy AI improves your decision-making processes