Review Note
Last Update: 11/26/2024 11:19 AM
Current Deck: Robotics and AI::Year 2::Term 1::Intro to ML::Week 5 - Multiclass Classification and Softmax Regression
PublishedCurrently Published Content
Text
In adaptive methods, each parameter update is scaled by {{c1::\( \frac{\alpha}{\sqrt{s_t} + \epsilon} \),}} where \( s_t \) is the accumulated gradient magnitude.
Extra
No published tags.
Pending Suggestions
No pending suggestions for this note.