When it comes to machine learning, understanding the math behind the models can be a game-changer. But is it really necessary to delve deep into the world of derivations and proofs to gain a deeper understanding of ML? I’ve been wondering if a mathematically rigorous approach to machine learning is essential for building a more holistic intuition around how ML works.
I’m not interested in applying these skills in a practical sense, but rather, I want to understand ML in a more profound way. With a solid grasp of calculus and linear algebra, I’m curious to know if a proof and derivation-heavy approach is necessary to achieve this deeper understanding.
ETH Zurich’s MSc Data Science program, with its emphasis on mathematical rigor, has piqued my interest. But I’m not convinced that this approach is the only way to gain a deeper understanding of ML. So, I’m turning to the community for insights.
Have you found that a mathematically rigorous approach to machine learning has helped you better understand the underlying concepts? Or do you think there are other ways to achieve a deeper understanding of ML?