Introducing Data-Driven Materials Informatics into Undergraduate Courses through a Polymer Science Workshop

Area of Science:

• Materials Science
• Polymer Engineering
• Computational Science

Background:

• Artificial intelligence (AI) and machine learning (ML) are increasingly vital in scientific research, including materials discovery and data analysis.
• Integrating AI/ML tools into research requires foundational understanding, which can be fostered through undergraduate education.

Purpose of the Study:

• To introduce fundamental machine learning concepts to undergraduate students.
• To enhance the adoption of AI/ML approaches in scientific research through early education.
• To integrate ML into a polymer science and engineering course via a practical workshop.

Main Methods:

• A hands-on, application-focused workshop was developed and delivered to undergraduate students in a polymer science and engineering course.
• Students engaged with the complete machine learning workflow: data cleaning, model training, performance evaluation, and result interpretation.
• A polymer solubility dataset, generated via visual inspection, was utilized for practical application.

Main Results:

• Pre- and post-workshop surveys demonstrated a measurable improvement in students' understanding of machine learning concepts.
• Student confidence in applying machine learning techniques to materials science problems increased significantly.
• The workshop successfully integrated new computational concepts into existing materials science coursework.

Conclusions:

• Foundational exposure to machine learning in undergraduate education can enhance its adoption in scientific research.
• A hands-on workshop is an effective method for teaching machine learning workflows to polymer science students.
• Integrating machine learning into materials science curricula bridges fundamental concepts with practical research applications.