Machine Learning Foundations and Applications in the Earth Systems Sciences is a series of modules designed to teach students core machine learning reasoning skills without requiring prerequisite programming knowledge. Machine learning tools and outputs are increasingly more popular in the Earth Systems Science workforce, thus, students should be prepared to interact with them upon degree completion. Additionally, advancements in the technology will warrant greater reliance on the responsible usage of pre-existing machine learning models and the interpretation of their outputs rather than the development of new models. This series of modules prepares students to be savvy users of machine learning tools by building their conceptual understanding of machine learning systems, encouraging critical scrutiny of data, and fostering judgment and decision making skills.

The first module is piloted in an advanced synoptic meteorology class at the Metropolitan State University of Denver. This module is designed to guide students through the very basics of supervised machine learning in the Earth Systems Sciences using a systems-thinking approach. They will discover how machine learning is used by scientists, the generalized process for model development, how data plays a crucial role in making good predictions, and how to be an effective and ethical user of machine learning tools. They also learn that machine learning is not a catch-all solution to every problem. Through simple schematics and graphs, students are guided through the conceptual process for developing and using supervised machine learning for science.

This poster will demonstrate the no-code approach to understanding supervised machine learning, reflect on the pilot session, and share lessons learned for future projects.