The course introduces concepts, techniques, and applications of machine learning techniques for geospatial research. It is organized into two related parts. The first part generally reviews the current state of the emerging field of geospatial data science and provides a survey of some advanced technology and tools to handle geospatial big data. The second part examines some popular machine learning techniques, such as supervised, unsupervised, and reinforcement learning. Students are expected to develop a good understanding of each machine learning method and how it works computationally. Students are then given the opportunity to practice these techniques to answer geographic research questions with computing tools. At the same time, students are required to review and critique the literature throughout the semester in the form of a research seminar.

• Introduction to geospatial data science (Week 1)

• Foundational and core concepts of machine learning, including automated analytical data modeling, pattern identification, and decision-making (Week 2)

• Technologies for geospatial big data processing (Weeks 3-5)

• Machine learning techniques and geospatial applications (Weeks 6-12)
1. Supervised learning, semi-supervised learning, and unsupervised learning
2. Decision trees and random forests
3. Support vector machines
4. Bayesian inference and estimation
5. Classification and regression trees
6. Deep learning

Applications of machine learning involving geospatial data, such as location-based social media data, GIS data, remote sensing imagery, and lidar acquired from satellite, airborne, unmanned aerial systems, and ground-based mobile devices and sensors. 

• Students will work individually or in teams to conduct projects using one or more machine learning techniques for 
applications demonstrating systems that learn from data and identify spatial-temporal patterns that can be used in decision-making with minimal human intervention (Weeks 13-14)

• Student presentations of their project objectives, workflows, results, and outcomes, with a critical assessment of machine 
learning data analysis (Week 15)

Expected Learning Outcomes
After completing the coursework, students are expected to:
1. develop a good understanding and critical insights on the emerging field of geospatial data science;
2. understand how some popular machine learning techniques work and critically assess how they were applied in published research projects; and
3. gain experience in teamwork, cooperation, and management to complete a research project using one or more of the machine learning techniques covered in this class.