Most of us have witnessed the accelerating impact of artificial intelligence (AI) on our lives. Few of us are not consuming AI in one form or another. Based on AI systems’ recommendations, we chose our jobs, partners, and the following video. The core technology behind most current AI systems is machine learning (ML).
A consequence of the recent AI hype is an increasing demand for accessible introductions to ML concepts. One challenge in providing high-quality introductions to ML is the audience’s diversity and needs.
Right from my start on the tenure track at Aalto University, I was faced with serving rapidly growing student cohorts in my ML courses. To this end, I have developed a simple three-component picture of ML. This picture simplifies and structures the learning process for students from diverse fields.
The three fundamental components of ML are
- Data Representation
- Model Design
- Loss Functions
We can decompose any ML method into these three components. Linear regression is obtained for one specific design choice for data, model and loss. Using a different design choice for the model results in deep learning methods. Reinforcement learning methods are characterized by using reward signals to construect loss functions.
The principles of this framework are extensively discussed in my textbook:
Machine Learning: The Basics
A. Jung, Springer, 2022.
Evidence-Based Approach
- Strong emphasis on interactive learning using online quizzes, coding assignments, and project-based assessments.
- Focus on conceptual clarity with practical exercises, such as Python notebooks hosted on JupyterHub.
Key Principles
- Keep lectures concise and targeted to specific learning goals.
- Provide curated resources, including textbooks (Machine Learning: The Basics), YouTube playlists, and GitHub repositories.
- Foster a feedback-adaptive teaching cycle while maintaining a consistent course schedule.
Student and Peer Recognition
- Teacher of the Year 2018, Department of Computer Science, Aalto University.
- Consistently high student ratings in teaching evaluations.
- Invited speaker and lecturer at multiple international institutions and conferences.