AI for geohazards
Course Overview
This course offers an introductory exploration of AI methods in Geohazards research, specifically landslide susceptibility mapping, providing a general overview of Geohazards, slope stability, and fundamentals of susceptibility mapping. We will find out how AI can complement traditional scientific methods and learn how to use machine learning for landslide susceptibility mapping, including training dataset composition, feature selection, and handling uncertainties in datasets. The course discusses the generation of time-dependent susceptibility maps and covers topics on map validation and also alternative methods for susceptibility mapping, beyond machine learning.
Details
Lecturer
Overview
This course offers an introductory exploration of AI methods in Geohazards research, specifically landslide susceptibility mapping, providing a general overview of Geohazards, slope stability, and fundamentals of susceptibility mapping. We will find out how AI can complement traditional scientific methods and learn how to use machine learning for landslide susceptibility mapping, including training dataset composition, feature selection, and handling uncertainties in datasets. The course discusses the generation of time-dependent susceptibility maps and covers topics on map validation and also alternative methods for susceptibility mapping, beyond machine learning.
Geohazards I
Get an introduction to the realm of geohazards, explore their global impact and the motivation driving research in this field.
Geohazards II
Discover the differences between hazard, susceptibility and risk and find out about secondary hazards and multi-hazard scenarios.
Slope Stability I
Introduction to the mechanisms involved in slope stability using infinite slope analysis and the factor of safety.
Slope Stability II
We look at how the physical description of slope stability can be used to draw conclusions about information for the data-driven analysis of slope stability.
Physics-based, data-driven and hybrid approaches
Over the decades, a wide range of methods for susceptibility and hazard assessment have been developed. In this video we will take a look at different approaches and how they can be applied to landslide susceptibility assessment.
Introduction to AI based Geohazards investigation
We will discuss machine learning in the context of Geohazards research in this video. We will take a first look at using machine learning for landslide susceptibility mapping.
Hazard Mapping I
We introduce the concept of susceptibility, hazard and hazard risk maps and what to consider when creating such a map.
Hazard Mapping II
In this video we take a look at landslide databases. They are essential for producing all kinds of hazard-related maps.
AI for Hazard Mapping I
We will discuss uncertainties and errors in datasets and processes used in creating training data
AI for Hazard Mapping II
We are going to discuss the composition of the training dataset for Random Forest classifier based landslide susceptibility mapping.
AI for Hazard Mapping III
We will discuss the features used in machine learning based landslide hazard mapping.
AI for Hazard Mapping IV
This video will give you an idea of the benefits, but also the challenges, of time-dependent susceptibility mapping based on machine learning.
Dynamic Features
We will discuss the incorporation of time-dependent features in the otherwise static susceptibility mapping process.
Validation
Validation plays a crucial role in scientific studies, encompassing a thorough examination of scientific consistency, uncertainty, accuracy, and errors to support confidence in the study. In this video, we delve into the validation process of a landslide susceptibility map.
