What is a neural network?
a) A biological system used for human cognition
b) A computational model inspired by the brain’s structure and functions
c) A method of solving mathematical problems
d) A process of optimizing mathematical equations
Which of the following is a common activation function in neural networks?
a) ReLU (Rectified Linear Unit)
b) Fibonacci sequence
c) Logistic regression
d) Gaussian distribution
What does a deep neural network have that distinguishes it from a simple neural network?
a) Fewer neurons
b) More layers between input and output
c) Faster processing speed
d) A higher accuracy in all cases
Which of the following is used to prevent overfitting in deep learning models?
a) Decreasing the size of the dataset
b) Using dropout regularization
c) Increasing the learning rate
d) Reducing the number of layers
What is backpropagation in neural networks?
a) A method of propagating errors backward through the network to update weights
b) A technique used to scale the input data
c) A function used to train a neural network’s final output
d) A process of reducing the complexity of the network
What is a common application of deep learning?
a) Solving basic algebra equations
b) Image and speech recognition
c) Sorting numbers in ascending order
d) Simple decision trees
What type of learning does a neural network perform?
a) Unsupervised learning
b) Supervised learning
c) Reinforcement learning
d) All of the above
Which of the following is a major disadvantage of deep neural networks?
a) High computational cost and time
b) Low accuracy
c) They cannot handle structured data
d) They only work with images
What does the “gradient” in gradient descent represent?
a) The value of the learning rate
b) The change in the cost function with respect to the weights
c) The input data fed into the network
d) The output prediction of the model
What is a convolutional neural network (CNN) primarily used for?
a) Sequence data analysis
b) Image processing and classification
c) Text classification
d) Time series prediction
What is the purpose of the pooling layer in a CNN?
a) To increase the number of features
b) To reduce the spatial dimensions of the input
c) To process sequential data
d) To normalize the input data
What does a recurrent neural network (RNN) excel at processing?
a) Static images
b) Temporal or sequential data
c) Structured tabular data
d) Linear regression tasks
What is the function of an activation function in a neural network?
a) To compute the final output of the network
b) To introduce non-linearity in the network
c) To initialize the weights
d) To normalize the input data
Which neural network architecture is commonly used in natural language processing tasks?
a) Convolutional Neural Networks (CNN)
b) Recurrent Neural Networks (RNN)
c) Radial Basis Function Networks (RBF)
d) Generative Adversarial Networks (GAN)
What does the term “overfitting” refer to in deep learning?
a) When the model is too simple to capture the underlying patterns
b) When the model performs well on the training data but poorly on new, unseen data
c) When the model cannot handle large datasets
d) When the model is undertrained
What is the role of the “bias” term in a neural network?
a) It adjusts the weights of the network
b) It shifts the activation function to allow the model to better fit data
c) It helps regularize the network
d) It initializes the output layer
Which optimization algorithm is commonly used to minimize the loss function in deep learning?
a) Newton’s method
b) Genetic algorithm
c) Stochastic Gradient Descent (SGD)
d) Particle Swarm Optimization
What is the purpose of the dropout technique in deep learning?
a) To improve the speed of computation
b) To randomly disable neurons during training to prevent overfitting
c) To increase the number of layers in a neural network
d) To reduce the complexity of the data
Which of the following is a key component of a generative adversarial network (GAN)?
a) Encoder and decoder
b) A generator and a discriminator
c) A feedback loop
d) A hidden layer
What is the purpose of “batch normalization” in deep learning?
a) To speed up training by normalizing the data in batches
b) To reduce the size of the network
c) To prevent overfitting by adding noise
d) To ensure that the model only uses batch data
Which of the following is an example of unsupervised learning?
a) Predicting house prices
b) Image segmentation
c) Image classification
d) Stock price prediction
What is the role of the “learning rate” in deep learning?
a) To control the speed at which the model learns by adjusting the weights
b) To control the number of layers in a neural network
c) To adjust the size of the input data
d) To determine the size of the output data
What does “vanishing gradient” refer to in deep learning?
a) When the gradients become too large and cause unstable learning
b) When the model has too many neurons in the output layer
c) When gradients become very small and stop the learning process in deep networks
d) When the model overfits the training data
Which of the following is a type of deep learning architecture that learns to represent data hierarchically?
a) Convolutional Neural Network (CNN)
b) Recurrent Neural Network (RNN)
c) Transformer
d) All of the above
What is the main function of an encoder-decoder architecture in neural networks?
a) To classify data into predefined categories
b) To generate output sequences from input sequences
c) To minimize the error rate of a network
d) To reduce the number of training examples needed
What does the term “underfitting” mean in machine learning?
a) The model captures too much noise from the data
b) The model is too simple and fails to capture the underlying patterns
c) The model is overfitted to the training data
d) The model performs perfectly on the training data
Which neural network technique is used to classify images in the computer vision domain?
a) Convolutional Neural Network (CNN)
b) Recurrent Neural Network (RNN)
c) Multilayer Perceptron (MLP)
d) Support Vector Machine (SVM)
What does the “softmax” function do in a neural network?
a) Converts predictions into probabilities that sum to 1
b) Calculates the error between predicted and actual values
c) Reduces the dimensionality of data
d) Ensures that the weights are adjusted correctly
What is “reinforcement learning”?
a) A type of learning where the model is trained with labeled data
b) A learning process based on rewards and penalties for actions
c) A supervised learning method for regression tasks
d) A technique for generating new data samples
