Skip to content
ml-recall

#sparse-features

3 questions from Optimization & Calculus, Classification

Timed test

3 questions · 3 min · score and explanations at the end

Take a test
Q 1 / 3Score 0 / 0

Question 1: Single choice

mediumAI-checkedOptimization & Calculus

AdaGrad (η=0.1\eta = 0.1, negligible ε\varepsilon) trains two weights. Weight A belongs to a frequent feature and has received the gradient 0.50.5 at each of the first 16 steps. Weight B belongs to a rare feature: its gradient was 0 at steps 1–15 and 0.50.5 at step 16. At step 16, how many times larger is B's step than A's?

Choose one option