Machine Learning

As a teaching assistant for undergraduate machine learning, I worked closely with students across the full learning loop: understanding concepts, implementing them in code, debugging assignments, and developing intuition for why algorithms behave the way they do.

My role went beyond grading. I helped translate lecture material into hands-on learning experiences, especially for students who were seeing the subject for the first time and needed support moving from formulas to actual model-building.

The work included:

• supporting lectures and clarifying foundational topics such as supervised learning, regularization, optimization, and evaluation
• preparing tutorial material and walkthroughs for assignments and lab sessions
• conducting hands-on sessions where students implemented models and analyzed results
• helping design and grade practical exercises that reinforced core concepts
• mentoring students during office hours and project discussions

One of the most rewarding parts of the role was helping students build intuition around model behavior instead of treating machine learning as a black box. A recurring focus was on questions such as:

• how to choose the right baseline for a task
• how to interpret bias-variance trade-offs in practice
• how to reason about train/test performance gaps
• how to debug feature choices, data leakage, and evaluation mistakes

Because the cohort was large, the role also required clear communication and consistency. I tried to make the course feel approachable by breaking down technical ideas into manageable pieces and being available to students outside class when they got stuck.

Overall, the experience strengthened both my teaching style and my own understanding of core ML fundamentals, especially the challenge of making technically rigorous material accessible to a broad student audience.