Subgradient Descent MCQ

Test your knowledge with important Subgradient Descent MCQ and their applications. These MCQs are beneficial for competitive exams too. Explore 30 + more Subgradient Descent MCQs at Bissoy

• Which of the following statements is not true about Subgradient method?
  ক

  Its convergence can be very fast

  খ

  It handles general non-differentiable convex problem

  গ

  It has no good stopping criterion

  ঘ

  It often leads to very simple algorithms
• Subgradient descent can be used at points where derivative is not defined.
  ক

  True

  খ

  False
• The Subgradient is a descent method.
  ক

  True

  খ

  False
• The step size rules in Subgradient are determined before the algorithm is run.
  ক

  False

  খ

  True
• Which of the following objective functions is not solved by Subgradient?
  ক

  Hinge loss

  খ

  L1 norm

  গ

  Perceptron loss

  ঘ

  TanH function
• In SVM problems, we cannot directly apply gradient descent but we can apply Subgradient descent.
  ক

  True

  খ

  False
• Step size, αk = α is a positive constant, independent of k is represented by Constant step size rule.
  ক

  True

  খ

  False
• Which of the following statements is not true about the Subgradient method?
  ক

  It has small memory requirement than interior-point methods

  খ

  It can be used for extremely large problems

  গ

  Simple distributed algorithm can be generated by combining sub gradient with primal or dual composition techniques

  ঘ

  It is much faster than Newton’s method in the unconstrained case
• Subgradient methods can be much slower than interior-point methods.
  ক

  True

  খ

  False
• Which of the following statements is not true about Subgradient?
  ক

  The step lengths are chosen via a line search

  খ

  It can be directly applied to non-differentiable functions

  গ

  It is an iterative method

  ঘ

  The step lengths are fixed ahead of time
• The Subgradient method is an algorithm for maximizing a non-differentiable convex function.
  ক

  True

  খ

  False