Using a decision tree to predict the number of COVID cases: a tutorial for beginners.

Area of Science:

• Biomedical Informatics
• Computational Biology
• Educational Technology

Background:

• Machine learning (ML) is crucial for advancing biomedical research, aiding in disease diagnosis, treatment, and prevention.
• The COVID-19 pandemic highlighted the utility of ML in public health predictions.
• Educating future researchers and clinicians in ML is essential for its continued application in science.

Approach:

• Developed an interactive learning module on ML and decision trees within the NIGMS Sandbox for Cloud-based Learning platform.
• Utilized a COVID-19 dataset, video tutorials, and annotated code for practical application.
• Leveraged cloud platforms like Google Colab and Google Cloud Platform (GCP) to eliminate local setup requirements.

Key Points:

• The module focuses on the fundamentals of decision trees and supervised machine learning concepts.
• Cloud-based access democratizes ML education by removing barriers to computational resource setup.
• Interactive learning with real-world data, like the COVID-19 dataset, enhances engagement.

Conclusions:

• The learning module effectively introduces machine learning basics and decision tree concepts to biology students.
• Cloud-based platforms facilitate accessible and time-efficient ML education.
• This approach has shown potential in increasing student interest in machine learning for research.