3 credits
Spring 2026 Lecture Upper Division
This course utilizes in-class activities personalized to each student's research interests. Our pedagogy is built on active engagement with complex problems. Hands-on projects are central to this experience, where you will apply machine learning (ML) to solve scientific and engineering challenges. The learning experience is organized into three phases: Phase 1: Foundations. Master the fundamentals of machine learning (supervised, unsupervised) and essential computational tools through hands-on case studies. Phase 2: Applications. Select and develop a research project using advanced SciML concepts like PINNs, autoencoders, or surrogate models, emphasizing practical problem-solving and collaboration. Phase 3: Communication. Produce a concise, publication-style research paper summarizing your project outcomes.
Learning Outcomes1Justify the selection and design of machine learning models for specific scientific and engineering problems, weighing their theoretical strengths and practical limitations against problem constraints.
2Implement robust and reproducible Scientific Machine Learning workflows, from data preprocessing to model validation, using professional coding standards and documentation.
3Critique the performance, uncertainty, and potential biases of machine learning models in scientific contexts and propose strategies for model improvement.
4Communicate complex computational research findings effectively to a specialized audience through both a formal written paper and an oral conference-style presentation.
Restrictions
made by Purdue students.