FEM 2017 Short Course: Advanced data analytics for mineral explorers

30.10.2017, 15:00 - varaukset
Kartta latautuu...

30.10.2017 - 31.10.2017
15:00 - 16:00

Hotel Hullu Poro

Welcome to FEM 2017 Pre-Conference Short Course!

Prof. Eun-Jung Holden and Tom Horrocks, PhD candidate, The University of Western Australia

Goal: The focus of the workshop is on concepts and applications of image analysis, machine learning and visualization for datasets used by mineral explorers

Course description: Principles of image analysis, machine learning and visualization. Human driven computer assisted geological interpretation. Data analytics methods and tools for diverse datasets including regional scale spatial data, outcrop images, downhole data and visualizing geophysical inversions.

Details of content: There has been an increasing interest by the minerals industry to capitalize on recent advances in data science to maximize their return on investment on collecting diverse geoscientific datasets throughout exploration, extraction and processing. The biggest challenge for the mineral explorers (greenfield or brownfield) is addressing uncertainty in geological interpretation. Understanding complex geology using sparse and diverse observations (often at varying scales) is not a trivial task, where human biases play a key role which result in highly inconsistent outcomes amongst and even within individuals. Computational algorithms can assist geological knowledge discovery through various analytical steps such as recognizing patterns of interest through signal/image processing, machine learning or statistical methods. However the geological insights by an interpreter, albeit inconsistent, which contribute to geologically feasible interpretation outcomes are hard to model for computational algorithms, especially considering highly variable existing knowledge, diverse and complex geological settings, and availability of different types of data at different scales.

This workshop will deliver an overview of fundamental concepts on image analysis, machine learning and data visualization methods, and their applications while ensuring the data analytical methods are interpretable and geologists’ insights can be incorporated into the data analysis workflow. Interpretation support methods and tools will be presented, including spatial exploration data from diverse sources for structural interpretation; and drill hole data for lithology modeling specifically for the Kevitsa Ni-Cu-PGE deposit. There will be hands-on exercise on Integrated Exploration Platform, a GIS based data analytics platform for ArcGIS which was developed in collaboration with the Geological Survey of Western Australia.

For more information, please visit fem.lappi.fi/fem-2017-short-courses