Teaching

Number of Credits
3

Course Designations and Attributes
LAS – Intermediate, 50% Graduate Coursework

Course Description
Introduces fundamental machine learning techniques for numerical modeling and data analysis and modern computer programming tools used to analyze, prepare, and visualize data from
common formats of datasets in the field of Earth and environmental sciences. Emphasizes opportunities to consider real-world applications for concepts in environmental data science.

Requisites
STAT 240, 301,324, 371 or Graduate/Professional Standing

Meeting Time and Location
TTh 2:30-3:45 pm @ 594 Russell Labs

Instructional Modality
In-person

How Credit Hours Are Met by the Course
This class meets for two, 75-minute class periods each week over the spring semester (3 hours per week, 42 hours in total). The students are expected to work on course learning, reading
course materials, completing homework tasks, and practicing activities for about 6-7 hours outside the classroom every week, 93 hours in total.

Regular and Substantive Student-Instructor Interaction
A qualified instructor will interact regularly and substantively with students through direct instruction during face-to-face class meetings twice a week and through personalized feedback on the weekly homework assignments. Students can ask questions anytime during class and after class by appointment.

Instructor: Min Chen, Assistant Professor
Instructor Availability: by appointment
Instructor Email: min.chen@wisc.edu
Teaching Assistant: N/A

Course Learning Outcomes
1. Demonstrate introductory skills in using collaboration technology (e.g. Jupyter Notebooks) to write, edit, and run programs in a scientific programming language (e.g. Python);
2. Recognize, read, write, and use common environmental dataset formats;
3. Use a scientific programming language (e.g. Python) to read and process environmental data;
4. Produce visualizations of environmental data, including basic scientific charts, statistics, and maps;
5. Understand the fundamentals of modern machine learning algorithms and gain experience with practical use of them;
6. Solve real-world data science problems individually and in teams;
7. Identify the frontiers in real-world environmental science challenges and how data science can help;
8. Identify a problem in environmental science that may be solved or better understood through data science, and provide a basic visualization or analysis of a data set associated
with that problem; (Undergraduate-only)
9. Develop in-depth spatial and/temporal analyses using advanced data science tools such as machine learning, visualize datasets related to your own research or anticipated area of research that meets the standards of scientific journals, critically evaluate your findings, and situate them within the larger context of current research literature; (Graduate-only)

Class Schedule

Week 1. Course overview; Setting up the programming environment

Week 2. Fundamentals of Python
Week 3. Fundamentals of Python, cont.
Week 4. Python Scientific packages
Week 5. Data visualization
Week 6. Analyzing spatial data
Week 7. Time series analysis
Week 8. Spring recess, No class
Week 9. Fundamentals of Machine Learning
Week 10. Machine learning packages in Python

Week 11. Case study 1: time series forecasting
Week 12. Case study 2：decision trees and random forests
Week 13. Case study 3: neural networks and deep learning
Week 14. Guest lecture
Week 15. Course project discussion; Questions and Answer sessions
Week 16. Final project due

Number of Credits: 3.0

Instructor: Phil Townsend

Co-Instructor: Natalie Queally, Eric Kruger, and Min Chen

Course Description: Introduction to major abiotic and biotic factors that influence forest ecosystem composition, structure, and function. Reviews important processes that influence the structure and function of forest ecosystems. Uses basic ecosystem concepts to elucidate the influence of anthropogenic(including forest management) and natural disturbances on forest ecosystem structure and function.