Introduction to data science for geologists





Introduction to data science for geologists

Martin Blouin (Geolearn)

Leaders and affiliations : M. Blouin (Geolearn), L. Perozzi (Geolearn), J. Simon (Geolearn), M. Claprood (INRS)

Description : Machine Learning is now everywhere in our lives. With some great applications in the Oil and Gas Industry and Mineral Exploration, it is now increasingly used in the geoscience industry. However, interpretability of the results it produces is a key point that generates apprehension with geoscientists. In this one-day course, we review the basic theoretical concepts behind machine learning, outline the history of its developments and the major algorithms relevant to the different applications in geoscience. We present applications in geosciences, the upsides of these methods and the major pitfalls to avoid. Through practical exercises, participants are introduced to unsupervised clustering algorithms (e.g. K-Means), dimensionality reduction techniques (e.g PCA) and supervised methods like random forest and neural networks. No prior coding or data science knowledge is required, but a strong interest in either statistics, modelling or data analysis is recommended. Exercises will be completed using a user-friendly and intuitive interface for data mining and Machine Learning. Attendees will need their own laptop.

Number of participants (Min/Max) : 10/50

Language : english

Duration : 1 day

Date : Pre-Congress

Associated special session / Field trip :

Special session: Artificial Intelligence and data management in mining exploration