Kernels, Data & Physics
This provides a theoretical framework for understanding complex ML problems, but it is incremental as it builds on existing NTK concepts without introducing new methods.
The lecture notes present the Neural Tangent Kernel (NTK) approach to address unsolvable problems in machine learning by reformulating them into tractable kernel formulations, with a focus on practical applications like data distillation and adversarial robustness.
Lecture notes from the course given by Professor Julia Kempe at the summer school "Statistical physics of Machine Learning" in Les Houches. The notes discuss the so-called NTK approach to problems in machine learning, which consists of gaining an understanding of generally unsolvable problems by finding a tractable kernel formulation. The notes are mainly focused on practical applications such as data distillation and adversarial robustness, examples of inductive bias are also discussed.