Education
Class Curriculum Vision
We aim to develop educational materials that will contain an à-la-carte menu of lectures, tutorials, and a collection of research-style course projects. What little existing educational materials for machine learning in geoscience are available are often high level and out of reach for beginners. Creating college level content suitable for individuals with a wider range of experience will benefit the entire geoscience community.
Look for these badges to identify books that make up the GeoSMART library. The GeoSMART library includes both class curriculum content and end to end machine learning workflows intended for a more experienced audience. If you'd like to contribute to the GeoSMART library, find more information here.
Materials
Machine Learning Curriculum
Data science for earth and planetary systems (ESS 469/569)
This UW course supported by GeoSMART serves as an introduction to data science for Earth and planetary systems. Class materials take the form of Jupyter Notebook tutorials with sample datasets, recorded lectures, links to Cloud computing resources and tutorials supported by the University of Washington eScience Institute. The GeoSMART classroom curriculum taught at the University of Washington consolidates these materials in a 10-12 week, four part multi-credit course covering:
- basics in software carpentry and access to cloud resources for 1-2 weeks
- introduction to ML tools for geosciences on small data sets for 3 weeks
- development and deployment of these tools to larger datasets as part of a course projects for 6-10 weeks
- project reporting with presentations, best practices for sharing open-source software, and tutorial development
Materials
Supporting Curricula
Python for research in water, weather, and climate (HWRS 401/501)
Data handling is a significant part of a machine learning workflow. This course is intended to introduce the basics of the modern data science stack for water, weather, and climate research, and can serve as a resource for learning data preparation for machine learning workflows. This course is based on the content originally developed for the University of Arizona class HWRS 401/501 “Tools for Data Handling and Analysis in Water, Weather, & Climate” also known as “HAS Tools”.
Geospatial Data Analysis with Python (CEE 467/CEWA 567)
Machine learning has been successfully used in remote sensing and geospatial applications. The Geospatial Data Analysis with Python class taught at the University of Washington explores geospatial data processing, analysis, interpretation, and visualization techniques using Python and open-source tools/libraries. The class content covers fundamental concepts, real-world data engineering problems, and data science applications using a variety of geospatial and remote sensing datasets that can inform data preparation for machine learning workflows. Highlights:
- aspects of both data engineering and data science, with exploratory data analysis approach
- learn how to programatically answer real-world remote sensing and GIS questions (and how to ask new questions)
- query and process geospatial data on-the-fly, without manual downloads
- limited emphasis on machine learning, but some examples scattered throughout labs (e.g., K-means clustering)
- examples focus on Washington state and Western U.S., with applications in snow hydrology and other CEE disciplines
Data Analysis in Water Sciences (CEE 465/CEWA 565)
This course covers fundamental topics related to data analysis using modern computer techniques, with applications to water sciences (but techniques are applicable to many science disciplines), including:
- statistical inference testing and error estimation and linear and quantile-based regression models
- monte Carlo simulation, time series analysis, bayes theorem, and data visualization
Future joint machine learning + geoscience curricula
Education
Class Curriculum Vision
We aim to develop educational materials that will contain an à-la-carte menu of lectures, tutorials, and a collection of research-style course projects. What little existing educational materials for machine learning in geoscience are available are often high level and out of reach for beginners. Creating college level content suitable for individuals with a wider range of experience will benefit the entire geoscience community.
Look for these badges to identify books that make up the GeoSMART library. The GeoSMART library includes both class curriculum content and end to end machine learning workflows intended for a more experienced audience. If you'd like to contribute to the GeoSMART library, find more information here.
Materials
Machine Learning Curriculum
Data science for earth and planetary systems (ESS 469/569)
This UW course supported by GeoSMART serves as an introduction to data science for Earth and planetary systems. Class materials take the form of Jupyter Notebook tutorials with sample datasets, recorded lectures, links to Cloud computing resources and tutorials supported by the University of Washington eScience Institute. The GeoSMART classroom curriculum taught at the University of Washington consolidates these materials in a 10-12 week, four part multi-credit course covering:
- basics in software carpentry and access to cloud resources for 1-2 weeks
- introduction to ML tools for geosciences on small data sets for 3 weeks
- development and deployment of these tools to larger datasets as part of a course projects for 6-10 weeks
- project reporting with presentations, best practices for sharing open-source software, and tutorial development
Materials
Supporting Curricula
Python for research in water, weather, and climate (HWRS 401/501)
Data handling is a significant part of a machine learning workflow. This course is intended to introduce the basics of the modern data science stack for water, weather, and climate research, and can serve as a resource for learning data preparation for machine learning workflows. This course is based on the content originally developed for the University of Arizona class HWRS 401/501 “Tools for Data Handling and Analysis in Water, Weather, & Climate” also known as “HAS Tools”.
Geospatial Data Analysis with Python (CEE 467/CEWA 567)
Machine learning has been successfully used in remote sensing and geospatial applications. The Geospatial Data Analysis with Python class taught at the University of Washington explores geospatial data processing, analysis, interpretation, and visualization techniques using Python and open-source tools/libraries. The class content covers fundamental concepts, real-world data engineering problems, and data science applications using a variety of geospatial and remote sensing datasets that can inform data preparation for machine learning workflows. Highlights:
- aspects of both data engineering and data science, with exploratory data analysis approach
- learn how to programatically answer real-world remote sensing and GIS questions (and how to ask new questions)
- query and process geospatial data on-the-fly, without manual downloads
- limited emphasis on machine learning, but some examples scattered throughout labs (e.g., K-means clustering)
- examples focus on Washington state and Western U.S., with applications in snow hydrology and other CEE disciplines
Data Analysis in Water Sciences (CEE 465/CEWA 565)
This course covers fundamental topics related to data analysis using modern computer techniques, with applications to water sciences (but techniques are applicable to many science disciplines), including:
- statistical inference testing and error estimation and linear and quantile-based regression models
- monte Carlo simulation, time series analysis, bayes theorem, and data visualization
Future joint machine learning + geoscience curricula