VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.
Machine Learning for Geobiology
Last Updated: 2026-02-05 16:07:14
Abstract
This class provides an introduction to machine learning concepts, techniques and algorithms and their applications in Geobiology. The course will cover both the fundamentals and application of machine learning techniques for geobiological research.
Objective
Students will learn the fundamentals of machine learning. An equal emphasis will be given to important geobiological discoveries achieved using machine learning methods. In completing the course, students will learn how to: - Generate hypotheses from data. - Make predictions from data. - Apply machine learning techniques to their own research
Content
In his exploration into the fundamental question of "What is life?", Schrödinger concluded that: "The unfolding of events in the life cycle of an organism exhibits an admirable regularity and orderliness, unrivalled by anything we meet with inanimate matter. We find it controlled by a supremely well-ordered group of atoms, which represent only a very small fraction of the sum total in every cell...To put it briefly, we witness the event that existing order displays the power of maintaining itself and of producing orderly events." Through the field of molecular biology, we now know these "supremely well-ordered groups of atoms" as DNA, RNA, and proteins and understand how they interact with one another to power life through the "Central Dogma of Molecular Biology" which involves the transcription of DNA to mRNA and translation of mRNA to proteins. Amazingly, all cellular organisms use these molecules composed of C, H, N, O, P and S in the same, orchestrated manner due to the fact that the instructions for chemical catalysis by RNA and proteins are encoded and stored in the DNA-based genomes of organisms. These instructions have been passed down from generation to generation and have been a central feature of life for over 3 billion years. While life converged on the "central dogma" relatively quickly, the mutability of the genome has enabled organisms to explore a wide variety of biological, physical, and chemical possibilities. The field of Evolutionary Biology provides a framework to study life's trajectory from origins to today and helps explain all of the biological complexity we observe. However, one must also remember that life does not operate in vacuum. Without the physical and chemical processes of our planet, life itself could not exist. Consequently, the physical, chemical and biological appearance of Earth are intricately entwined throughout Earth History. Geobiology is a field that studies the interactions between life and the environment. The central dogma of molecular biology and fundamental laws of physics and chemistry guide the interactions between the living and physical world. These interactions result in geobiological signals that can be detected throughout the geologic and genetic record. As geobiologists, our goal is to discover these signals through patterns in data, understand the processes that produce these patterns and make predictions about the conditions in which such patterns arise. Machine Learning can help us achieve these objectives and advance our field. This course will cover a variety of machine learning topics used by geobiologists, including: - Regression - Unsupervised learning (e.g. k-means clustering, PCA and t-SNE) - Hidden Markov Models - Ensemble learning methods (e.g. Random Forest) - Bayesian Inference - Convolutional Neural Networks
General Information
- Language
- English
- Levels
- MSC
- Frequency
- Yearly recurring
Examination
- Type
- graded semester performance
Course Components
| Type | Title | Time & Place | Hours |
|---|---|---|---|
| lecture | Machine Learning for Geobiology |
|
2 h weekly |
| practical/laboratory course | Machine Learning for Geobiology - Project | No time listed | 2 h weekly |
Offered In
-
-
Electives (Courses can be chosen from the complete offerings of the ETH Zurich and University of Zurich (according to prior agreement with the MSc Committee).)
-