061: Interpolation, Extrapolation and Linearisation (Prof. Yann LeCun, Dr. Randall Balestriero)
Machine Learning Street Talk (MLST)
Navigating Interpolation and Extrapolation in Machine Learning
This chapter examines the complexities of feature engineering with a focus on interpolation, using watch face pixel grids as a case study. It highlights the limitations of deep learning models in handling high-dimensional data and the necessity of strong inductive biases for better generalization. Through discussions on the manifold hypothesis and the curse of dimensionality, it underscores the challenges posed by traditional interpolation methods in machine learning.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
