Computer scienceData scienceMachine learningIntroduction to deep learningOptimizers

Gradient descent with momentum

The theta update

Report a typo

You are performing gradient descent with a learning rate (ϵ\epsilon) of 0.1 on a convex loss function. In the first iteration, the gradient is -3. What is the parameter update θt+1−θt\theta_{t+1} - \theta_{t} in this iteration?

Hint: Check the formula for θt+1−θt\theta_{t+1} - \theta_{t} from theory.

Select one option from the list

Create a free account to access the full topic