Technology
Algorithms combining Geostatistics and Machine Learning
By leveraging a geostatistical asset that has been continuously developed and applied for over 10 years in the natural resources industry, and enhancing it with Machine Learning, we pave the way for innovative local modeling of consumer interests and behaviors.
We cross-reference all relevant data and perform modeling of the variables of interest
Geostatistics provides a probabilistic framework in which the data of interest are considered as realizations of a set of possible realizations of a random spatial function. This function is supposed to represent the phenomenon under study. This framework allows the construction of local estimation models that minimize the estimation error.
Step 1 - Data Gathering
Your data (optional)
CRM, campaigns, surveys, etc.
Open data
INSEE databases, SIREN, etc.
Collected data
Social networks and large-scale web surveys.
Step 2 - Non-stationary Multivariable Geostatistical Modeling
10 years of industrial development and application, patented technology for determining local parameters.
Step 3 - GeoInsights
We develop innovative geostatistical models and ensure their optimal configuration
Innovation is at the heart of our development. Each year, we make significant investments to improve our local modeling algorithms.
Ever more precise modeling
One of our recent innovations combines modeling and learning tools from Machine Learning with the most advanced Geostatistics techniques (SPDE approaches).
True expertise in model implementation and configuration
Developing innovative models is one thing, applying them to real data is another. The experience and expertise of our team ensure optimal implementation and configuration of the models.
