Print this Page

Topical Session 2: Machine Learning for Geoscience Modelling

Conveners: V. Demyanov (Heriot-Watt University, UK), M. Kanevski (University of Lausanne, Switzerland)

The session aims to demonstrate how machine learning solves the present day challenges in geosciences applications.  Machine learning has gained the increased attention with more data becoming available to describe natural systems in the digital era, such as environmental and climate modelling, pollution and natural hazards mapping, subsurface reservoir characterisation and optimisation, etc.  Learning based techniques have proven their power in making prediction and gain understanding of the geoscience systems behaviour based on “big data” and knowledge integration. Machine learning methods are often seen as complementary/competitive to geostatistics in their flexibility in modelling data and knowledge integration. Many geosciences research groups across the world are using data driven algorithms for various applications. The session will invite multi-disciplinary contributions that will demonstrate recent advances in applying learning based algorithms to geosciences problems.

Permanent link to this article: http://www.iamg2018.org/index.php/topical-session-2-machine-learning-for-geoscience-modelling/