1. What is deep learning? a) A type of machine learning algorithm b) A branch of artificial intelligence c) A technique for training neural networks d) All of the above Answer: d) All of the above
2. Which of the following is not a popular deep learning framework?
a) TensorFlow b) PyTorch c) Keras d) Scikit-learn Answer: d) Scikit-learn
3. What is the purpose of an activation function in a neural network?
a) It introduces non-linearity to the network b) It determines the output of a neuron c) It helps in backpropagation d) All of the above Answer: d) All of the above
4. What is the role of gradient descent in deep learning?
a) To minimize the loss function b) To find the optimal weights for the neural network c) To update the network's parameters d) All of the above Answer: d) All of the above
5. What is the vanishing gradient problem?
a) When the gradients in a deep neural network become extremely small b) When the gradients in a deep neural network become extremely large c) When the gradients in a shallow neural network become extremely small d) When the gradients in a shallow neural network become extremely large
Answer: a) When the gradients in a deep neural network become extremely small
6. What is backpropagation used for in deep learning?
a) To calculate the gradients for updating the network's parameters b) To propagate errors from the output layer to the input layer c) To train a neural network d) All of the above Answer: d) All of the above
7. Which of the following is a common loss function used for binary classification in deep learning? a) Mean Absolute Error (MAE) b) Mean Squared Error (MSE) c) Binary Cross-Entropy d) Categorical Cross-Entropy Answer: c) Binary Cross-Entropy 8. What is the purpose of dropout regularization in deep learning? a) To reduce overfitting b) To increase the model's capacity c) To improve the training speed d) To handle imbalanced datasets Answer: a) To reduce overfitting 9. What is an epoch in deep learning? a) The number of layers in a neural network b) The number of training examples in a dataset c) The number of times the entire dataset is passed through the neural network during training d) The number of neurons in a layer Answer: c) The number of times the entire dataset is passed through the neural network during training 10. What is the purpose of a convolutional layer in a convolutional neural network (CNN)? a) To reduce the dimensionality of the input b) To extract spatial features from the input c) To classify the input data d) To apply non-linear transformations to the input Answer: b) To extract spatial features from the input 11. Which activation function is commonly used in the hidden layers of a deep neural network? a) ReLU (Rectified Linear Unit) b) Sigmoid c) Tanh d) Softmax Answer: a) ReLU (Rectified Linear Unit) 12. What is the purpose of pooling layers in a convolutional neural network (CNN)? a) To reduce the size of the input data b) To perform spatial downsampling c) To extract the most important features d) All of the above Answer: d) All of the above
13. Which of the following is a common optimization algorithm used in deep learning? a) Gradient Descent b) Stochastic Gradient Descent (SGD) c) Adam d) All of the above Answer: d) All of the above
14. What is the purpose of data normalization in deep learning?
a) To scale the input data to a fixed range b) To improve the convergence of the optimization algorithm c) To make the input data more interpretable d) To preprocess the data for visualization Answer: a) To scale the input data to a fixed range
15. What is the purpose of a recurrent neural network (RNN)?
a) To handle sequential data b) To classify images c) To perform dimensionality reduction d) To generate synthetic data Answer: a) To handle sequential data
16. Which type of RNN architecture is used to address the vanishing gradient problem? a) Long Short-Term Memory (LSTM) b) Gated Recurrent Unit (GRU) c) Simple RNN d) Bidirectional RNN Answer: a) Long Short-Term Memory (LSTM)
17. Which deep learning technique is used for generating new, realistic data samples? a) Generative Adversarial Networks (GANs) b) Convolutional Neural Networks (CNNs) c) Reinforcement Learning d) Transfer Learning Answer: a) Generative Adversarial Networks (GANs) 18. What is the purpose of transfer learning in deep learning? a) To reuse pre-trained models on new tasks b) To transfer knowledge from one domain to another c) To speed up the training process d) All of the above Answer: d) All of the above
19. What is the purpose of a loss function in deep learning?
a) To measure the difference between predicted and actual values b) To guide the training process c) To compute the gradients for updating the network's parameters d) All of the above Answer: d) All of the above
20. Which deep learning technique is used for sequence-to-sequence tasks, such as machine translation? a) Attention Mechanism b) Convolutional Neural Networks (CNNs) c) Transfer Learning d) Autoencoders Answer: a) Attention Mechanism
21. Which of the following is a common activation function used in the output layer for binary classification? a) Sigmoid b) ReLU (Rectified Linear Unit) c) Tanh d) Softmax Answer: a) Sigmoid 22. What is the purpose of batch normalization in deep learning? a) To reduce internal covariate shift b) To accelerate the training process c) To improve the generalization of the model d) All of the above Answer: d) All of the above
23. Which deep learning technique is used for unsupervised feature learning? a) Autoencoders b) Convolutional Neural Networks (CNNs) c) Recurrent Neural Networks (RNNs) d) Reinforcement Learning Answer: a) Autoencoders
24. What is the purpose of an embedding layer in deep learning?
a) To map high-dimensional input to a lower-dimensional space b) To convert categorical variables into numerical representations c) To learn distributed representations of words or entities d) All of the above Answer: d) All of the above
25. Which of the following is a popular deep learning architecture for object detection? a) YOLO (You Only Look Once) b) LSTM (Long Short-Term Memory) c) VGG (Visual Geometry Group) d) GAN (Generative Adversarial Network) Answer: a) YOLO (You Only Look Once)
26. What is the purpose of early stopping in deep learning?
a) To prevent overfitting b) To save training time c) To improve the model's generalization ability d) To avoid local optima in the loss function
Answer: a) To prevent overfitting
27. Which deep learning technique is used for learning latent representations from unlabelled data? a) Self-supervised learning b) Reinforcement Learning c) Transfer Learning d) Unsupervised Learning Answer: d) Unsupervised Learning
28. What is the purpose of dropout in deep learning?
a) To randomly disable neurons during training b) To reduce the model's capacity c) To regularize the network and prevent overfitting d) All of the above Answer: c) To regularize the network and prevent overfitting 29. Which of the following is a common metric used to evaluate the performance of a classification model in deep learning? a) Accuracy b) Mean Absolute Error (MAE) c) Mean Squared Error (MSE) d) R-squared Answer: a) Accuracy 30. Which deep learning technique is used for learning from delayed rewards? a) Reinforcement Learning b) Supervised Learning c) Unsupervised Learning d) Transfer Learning Answer: a) Reinforcement Learning 31. Which of the following deep learning architectures is commonly used for image segmentation? A. ResNet B. U-Net C. AlexNet D. GANs Answer: B) U-Net
32. Which type of image segmentation assigns each pixel in the image a class label?
A. Object detection B. Semantic segmentation C. Instance segmentation D. Supervised learning Answer: B) Semantic segmentation
33. Which popular word embedding method captures the semantic meaning of words by positioning similar words closely in vector space? A. Bag of Words (BoW) B. Term Frequency-Inverse Document Frequency (TF-IDF) C. Word2Vec D. One-Hot Encoding Answer: C) Word2Vec
34. In the Transformer model, what does "BERT" stand for?
A. Bidirectional Encoder Representations from Transformers B. Basic Encoder Representation Technique C. Bidirectional Embedding of Recurrent Transformers D. Binary Encoder Representation Technology Answer: A) Bidirectional Encoder Representations from Transformers 35. In 2021, Google Research introduced a model specifically designed for handling extremely large text datasets. This model is known as: A. GPT-Neo B. GShard C. Turing-NLG D. BigGAN Answer: B) GShard 36. Which type of ANN architecture is commonly used for image classification tasks? A. Convolutional Neural Network (CNN) B. Recurrent Neural Network (RNN) C. Radial Basis Function Network (RBFN) D. Multilayer Perceptron (MLP) Answer: A) Convolutional Neural Network (CNN)
37. Which deep learning technique is used for learning from delayed rewards? A. Reinforcement Learning B. Supervised Learning C. Unsupervised Learning D. Transfer Learning Answer: A) Reinforcement Learning
38. What are the general tasks that are performed with backpropagation algorithm? A. pattern mapping B. function approximation C. prediction D. all of the mentioned Answer: D) all of the mentioned 39. Which of the following statements is not true about Kernel methods? a) It works by embedding the input data to some high dimensional feature space b) Embedding into feature space can be determined uniquely by specifying a kernel function that computes the dot product between data points in the feature space c) It defines only the linear mapping to the feature space d) Expensive computations in the high dimensional feature space can be avoided by evaluating the kernel function Answer:c) It defines only the linear mapping to the feature space
40. What is the use of MLFFNN?
a) to realize structure of MLP b) to solve pattern classification problem c) to solve pattern mapping problem d) to realize an approximation to a MLP Answer:d) to realize an approximation to a MLP
41.What is the advantage of basis function over mutilayer feedforward neural networks? a) training of basis function is faster than MLFFNN b) training of basis function is slower than MLFFNN c) storing in basis function is faster than MLFFNN d) none of the mentioned Answer:a) training of basis function is faster than MLFFNN 42. What consist of a basic counterpropagation network? a) a feedforward network only b) a feedforward network with hidden layer c) two feedforward network with hidden layer d) none of the mentioned Answer:c) two feedforward network with hidden layer
43. How can false minima be reduced in case of error in recall in feedback neural networks? a) by providing additional units b) by using probabilistic update c) can be either probabilistic update or using additional units d) none of the mentioned Answer: b) by using probabilistic update
44. What property should a feedback network have, to make it useful for storing information? a) accretive behaviour b) interpolative behaviour c) both accretive and interpolative behaviour d) none of the mentioned Answer: a) accretive behaviour 45. What is the objective of a pattern storage task in a network? a) to store a given set of patterns b) to recall a give set of patterns c) both to store and recall d) none of the mentioned Answer: c) both to store and recall 46. Which is a simplest pattern recognition task in a feedback network? a) heteroassociation b) autoassociation c) can be hetero or autoassociation, depends on situation d) none of the mentioned Answer: b) autoassociation 47. What happens when number of available energy minima be less than number of patterns to be stored? a) pattern storage is not possible in that case b) pattern storage can be easily done c) pattern storage problem becomes hard problem for the network d) none of the mentioned Answer: c) pattern storage problem becomes hard problem for the network 48. What happens when number of available energy minima be more than number of patterns to be stored? a) no effect b) pattern storage is not possible in that case c) error in recall d) none of the mentioned Answer: c) error in recall
49. Why there is error in recall, when number of energy minima is more the required number of patterns to be stored? a) due to noise b) due to additional false maxima c) due to additional false minima d) none of the mentioned Answer: c) due to additional false minima 50. Generalization feature of a multilayer feedforward network depends on factors? a) architectural details b) learning rate parameter c) training samples d) all of the mentioned Answer: a) architectural details 51. What is accretive behaviour? a) not a type of pattern clustering task b) for small noise variations pattern lying closet to the desired pattern is recalled. c) for small noise variations noisy pattern having parameter adjusted according to noise variation is recalled d) none of the mentioned Answer: b) for small noise variations pattern lying closet to the desired pattern is recalled. 52. What is Interpolative behaviour? a) not a type of pattern clustering task b) for small noise variations pattern lying closet to the desired pattern is recalled. c) for small noise variations noisy pattern having parameter adjusted according to noise variation is recalled d) none of the mentioned Answer: c) for small noise variations noisy pattern having parameter adjusted according to noise variation is recalled 53. What is the feature that doesn’t belongs to pattern mapping in feeddorward neural networks? a) recall is direct b) delta rule learning c) non linear processing units d) two layers Answer: d) two layers 54. What is the primary purpose of a Recurrent Neural Network (RNN)? a) Image classification b) Text generation c) Reinforcement learning d) Object detection Answer: b) Text generation
55. Which layer type is typically used to capture sequential dependencies in an RNN? a) Input layer b) Hidden layer c) Output layer d) Activation layer Answer: b) Hidden layer
56. What is the advantage of using recurrent layers in an RNN?
a) They can capture temporal dependencies in the input data b) They can handle variable-length inputs c) They can generate synthetic data d) They can handle non-linear transformations Answer: a) They can capture temporal dependencies in the input data
57. What is the purpose of the hidden state in an RNN?
a) To store the information from the previous time step b) To adjust the learning rate during training c) To compute the gradients for backpropagation d) None of the above Answer: a) To store the information from the previous time step 58. Which activation function is commonly used in the recurrent layers of an RNN? a) ReLU (Rectified Linear Unit) b) Sigmoid c) Tanh (Hyperbolic Tangent) d) Softmax Answer: c) Tanh (Hyperbolic Tangent)
59. What is the purpose of the time step parameter in an RNN?
a) To determine the number of recurrent layers in the network b) To adjust the learning rate during training c) To specify the length of the input sequence d) None of the above Answer: c) To specify the length of the input sequence
60. Which layer type is commonly used to initialize the hidden state in an RNN? a) Input layer b) Hidden layer c) Output layer d) Activation layer Answer: b) Hidden layer
61. What is the purpose of the bidirectional RNN architecture?
a) To handle sequential data in both forward and backward directions b) To reduce the computational complexity of the network c) To adjust the learning rate during training d) None of the above Answer: a) To handle sequential data in both forward and backward directions 62. Which layer type is responsible for making final predictions in an RNN? a) Input layer b) Hidden layer c) Output layer d) Activation layer Answer: c) Output layer
63. What is the purpose of the recurrent connection in an RNN?
a) To propagate the hidden state across different time steps b) To adjust the weights and biases of the network c) To reduce the dimensionality of the input data d) None of the above Answer: a) To propagate the hidden state across different time steps
64. Which layer type is commonly used in RNNs for sequence-to-sequence tasks? a) Input layer b) Hidden layer c) Output layer d) Attention layer Answer: d) Attention layer
65. What is the purpose of the backpropagation through time (BPTT) algorithm in RNN training? a) To compute the gradients and update the network's parameters b) To adjust the learning rate during training c) To prevent overfitting by regularizing the model d) None of the above Answer: a) To compute the gradients and update the network's parameters 66. Which layer type is commonly used in RNNs to handle variable-length inputs? a) Input layer b) Hidden layer c) Output layer d) None of the above Answer: a) Input layer
67. What is the purpose of the initial hidden state in an RNN?
a) To provide the starting point for the recurrent computation b) To adjust the learning rate during training c) To compute the gradients for backpropagation d) None of the above Answer: a) To provide the starting point for the recurrent computation
68. Which layer type is responsible for handling the output at each time step in an RNN? a) Input layer b) Hidden layer c) Output layer d) Activation layer Answer: c) Output layer
69. What is the purpose of the teacher forcing technique in RNN training? a) To adjust the learning rate during training b) To propagate the gradients through time c) To reduce the computational complexity of the network d) None of the above Answer: b) To propagate the gradients through time 70. Which layer type is commonly used in RNNs for language modeling tasks? a) Input layer b) Hidden layer c) Output layer d) None of the above Answer: c) Output layer
71. What is the purpose of the sequence-to-vector architecture in an RNN?
a) To process an input sequence and produce a fixed-length representation b) To adjust the weights and biases of the network c) To reduce the dimensionality of the input data d) None of the above Answer: a) To process an input sequence and produce a fixed-length representation
72. Which layer type is responsible for introducing non-linearity in an RNN?
a) Input layer b) Hidden layer c) Output layer d) Activation layer Answer: d) Activation layer
73. What is the purpose of the forget gate in a Gated Recurrent Unit (GRU)? a) To control the flow of information from the previous hidden state b) To adjust the learning rate during training c) To compute the gradients for backpropagation d) None of the above Answer: a) To control the flow of information from the previous hidden state 74. Which layer type is commonly used in RNNs for machine translation tasks? a) Input layer b) Hidden layer c) Output layer d) Attention layer Answer: d) Attention layer
75. What is the purpose of the peephole connections in a Long Short-Term Memory (LSTM) network? a) To allow the cell state to influence the gating mechanisms b) To adjust the learning rate during training c) To introduce non-linearity to the network d) None of the above Answer: a) To allow the cell state to influence the gating mechanisms
76. Which layer type is responsible for handling variable-length outputs in an RNN? a) Input layer b) Hidden layer c) Output layer d) None of the above Answer: c) Output layer
77. What is the purpose of the cell state in an LSTM network?
a) To store long-term dependencies in the input sequence b) To adjust the learning rate during training c) To compute the gradients for backpropagation d) None of the above Answer: a) To store long-term dependencies in the input sequence 78. Which layer type is commonly used in RNNs for speech recognition tasks? a) Input layer b) Hidden layer c) Output layer d) None of the above Answer: c) Output layer
79. What is the purpose of the input gate in an LSTM network?
a) To control the flow of information from the current input b) To adjust the learning rate during training c) To introduce non-linearity to the network d) None of the above Answer: a) To control the flow of information from the current input
80. Which layer type is responsible for handling variable-length inputs and outputs in an RNN? a) Input layer b) Hidden layer c) Output layer d) None of the above Answer: d) None of the above
