Metacademy - Differential geometry for machine learning
web.archive.org
Metacademy - Differential geometry for machine learning
CNN Explainer
poloclub.github.io
If you could figure out what patterns of neural activity in a 100-dimension population are fundamental to that population – and which are just recycled combinations of those fundamental patterns – you could explain that neural population with fewer than 100 dimensions.
Grace Lindsay • Models of the Mind
This “blessing of nonuniformity,” whereby data is not spread uniformly in (hyper) space, is often what saves the day.
Pedro Domingos • The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World
Any understanding of the mundane world that we develop necessarily involves dimension reduction and simplification: so the story goes. We may end up with a tractable model that predicts reasonably well, but we’ll never get our error term to zero, and we mustn’t confuse our toy conceptions for a high-fidelity reproduction of “reality.”
eigenrobot • the map is of the territory
Essential Math for Data Science
