Quiz Space

Deep Learning · End Term · 24 Dec 2023 · September 2023 term · Set ADB3

Question 4: Suppose that a neural network has millions of parameters …

Question 4

+3 marksOne or more correct options

Suppose that a neural network has millions of parameters (weights and biases). A team decides to use an optimization algorithm with a learning rate scheme that is local to each parameter in the network. Moreover, the learning rate changes in each iteration such that it should decrease on the steep surface and increase on the gentle surface. Which of the following optimization algorithms satisfy the team’s requirements?

Select all that apply.

  1. A

    GD with an exponentially decaying learning rate scheduler

  2. B

    AdaGrad

  3. C

    AdaM

  4. D

    NADAM

  5. E

    RMSProp

  6. F

    SGD with line search

Show answer

Correct answers

  • C

    AdaM

  • D

    NADAM

  • E

    RMSProp

See the whole paperTake it as a mock test

Question 4 of 18 in the IIT Madras BS Deep Learning (Deep Learning) End Term paper sat on 24 Dec 2023, in the September 2023 term (IIT M DEGREE FN EXAM FDB1 24 Dec 2023). It carries 3 marks.

This question was also asked in

More questions from this paper

  1. Q1Figure question
  2. Q2Figure question
  3. Q3Figure question
  4. Q5Suppose we divide the available training samples into mini batches of size 32 to train a model with mini-batch gradient…
  5. Q6Select all correct statements about the bias and variance of deep learning models.
  6. Q7Figure question
  7. Q8Figure question
  8. Q9Figure question
  9. Q10How many neurons are there in the network? Every neuron in the network has bias associated with it?\ Note: A neuron is …
  10. Q11How many parameters (including bias) does the network have to learn? Assume dropout regularization is applied.
  11. Q12What is the size of the vocabulary |V |?
  12. Q13Suppose we consider the three words (wood,woodchuck, much). Assume we use one-hot encoded vector representation for all…
  13. Q14Suppose that we need to develop an RNN model for sentiment classification tasks. The input to the model is a sentence c…
  14. Q15Suppose that we need to develop an RNN model for sentiment classification tasks. The input to the model is a sentence c…
  15. Q16Suppose that we need to develop an RNN model for sentiment classification tasks. The input to the model is a sentence c…
  16. Q17Figure question
  17. Q18Consider an intermediate feature map H obtained after applying convolution operation on the input X using kernel F. H =…