CIM 2017: Demystifying artificial intelligence in exploration

By IOS team, April 28, 2017

Hypercube, Artificial intelligence, Machine learning, Predictive modelling, classification algorithms, Bayesian logic: Let’s demystify!

From April 29 to May 3rd, the Canadian Institute of Mining will hold its annual national convention in Montréal’s convention centre. This annual gathering is the great opportunity to learn about new trends, issues and innovation of every aspect of mining operations, from exploration to reclamation. Talking innovation… IOS invites you to a presentation about the use of artificial intelligence or machine learning in mineral endowment prediction and targeting. The talk (#1766) will be held on Tuesday, May 2nd, in the “Geology and Exploration” session.

The talk “Predictive modelling of regional mineral potential: The Hypercube approach” aims to provide a comprehensive introduction to Bayesian logic and how is it used to recognize the complex footprint of mineral deposits within large and heterogeneous geodatabase, from the mine to the regional scales. The subjacent principles will be explained in simple terms that geologists and engineers unfamiliar with concepts of data mining can understand, in a manner to challenge their skepticism and encourage them to see powerful tools. A case study will be presented, dealing with the prediction of orogenic gold occurrences in James Bay area, extracted from the public domain regional dataset of the SIGÉOM. How the system has been capable to rightfully predict 96% of the occurrence will be mathematically demonstrated.

See you there!


Predictive model for orogenic gold occurrences over the entire James Bay area (90,000 km2).  Coloured areas were selected as zone of favorability, while little black triangles are known gold occurrences.

< Back to news