6CSM1 B2 QUIZZ Optimization

enter your roll number

Which of the following is a common loss function used for binary classification in deep learning?

• Mean Absolute Error (MAE)

• Binary Cross-Entropy

• Mean Squared Error (MSE)

• Categorical Cross-Entropy

What is an epoch in deep learning?

• The number of layers in a neural network

• The number of training examples in a dataset

• The number of neurons in a layer

• The number of times the entire dataset is passed through the neural network during training

What is the purpose of pooling layers in a convolutional neural network (CNN)?

• To reduce the size of the input data

  
  

  
  

• To perform spatial downsampling

• To extract the most important features

• All of the above

What is the purpose of data normalization in deep learning?

• To improve the convergence of the optimization algorithm

• To make the input data more interpretable

• To scale the input data to a fixed range

  
  

• To preprocess the data for visualization

What is the purpose of early stopping in deep learning?

• To save training time

  
  

• To prevent overfitting

• To improve the model's generalization ability

• To avoid local optima in the loss function

What does a gradient descent algorithm do?

• NONE

• Both C and D

• Tries to find the parameters of a model that minimizes the cost function

• Adjusts the weights at the input layers

What is the main benefit of batch Gradient in deep learning over stochastic gradient?

• It provides more accurate estimate of the gradient

• It reduces memory usage during training

• It requires less computation

• It converges faster to the optimal solution

Which of the following optimizers is most commonly used in training deep neural networks?

• Stochastic Gradient Descent (SGD)
  

• Adam

• Gradient Descent with Momentum

• AdaGrad

Which optimizer is often preferred for sparse data (like text data in NLP)?

• Adam

• SGD

• RMSprop

• AdaGrad