
Dropout as Implicit Bagging in Deep Learning
March 7, 2026
A concrete way to understand dropout in deep learning: shared-parameter ensemble training that approximates bagging without training separate models.
Read postMachine learning theory notes on regularization, model structure, optimization, and the reasoning behind familiar techniques.
2 posts currently filed under ML Theory.

A concrete way to understand dropout in deep learning: shared-parameter ensemble training that approximates bagging without training separate models.
Read post
Chapter 6 sharpened how I think about architecture as a structural assumption, not just a tuning choice.
Read postCopyright © 2020 - 2026 Alex Leung