MACHINE LEARNING AND GEOCHEMICAL FOOTPRINTING: EXAMPLES FROM THE ABITIBI CLAY BELT.

By Team IOS, Monday February 12. 2018

A talk within a Short Course, PDAC, Toronto, March 2nd and 3rd 2018
http://www.pdac.ca/convention/programming/short-courses/sessions

 

If you think geophysics is complex? What about geochemistry which may involve up to 100 variables at every sampling site! The complexity of the natural processes leading to geochemical footprints of mineral deposits makes their prediction almost impossible for human mind, or it is so long and tedious to deceipherthat it is unaffordable for the mineral industry to do it properly. General practices involve statistics with various level of mathematical sophistication. However, a different approach has been developed by IOS team, based on Bayesian logics and machine learning. The method enables to define complex and variable footprints over heterogeneous targets, and to predict new occurrences from such signature.

 

The method will be presented as a talk within the short course « Exploration geochemistry: Fundamentals and case histories ». This workshop is organized by Mrs Beth McClenaghan and Mrs Lynda Bloom and will be held before PDAC, on the 2nd and 3rd of March in Toronto. The workshop program includes a dozen of talks from leading geochemists covering topics from surveys planning to case study reviews. Examples of application of our approach from soil and peat surveys within the Abitibi Clay Belt will be presented.

 

Upgrade you stay at PDAC: Improve your abilities and network with experts.

 

"There are those who follow the rules... and those who define them!"

 

Kaskoo

Example of a region in Northern Abitibi where the interpretation of peat geochemistry based on artificial intelligence allowed the prediction of mineralized occurrence under 50 metres of overburden deemed impermeable.

 

< Back to news