1. What is a neural network in deep learning?

A) A network of neurons in the human brain B) A computational model composed of layers of interconnected nodes that process data C) A statistical method for data analysis D) A type of data storage system

2. What does "backpropagation" refer to in deep learning?

A) The process of sending data backward through the network B) An algorithm for updating the weights in a neural network C) A method for data augmentation D) A technique for preventing overfitting

3. What is the purpose of an activation function in a neural network?

A) To introduce non-linearity into the network B) To activate neurons randomly C) To measure the error rate D) To initialize the weights

4. What is "overfitting" in the context of deep learning?

A) When a model performs well on training data but poorly on unseen data B) When a model has too few parameters C) When a model cannot converge during training D) When a model generalizes well to new data

5. Which technique can help prevent overfitting in a neural network?

A) Using dropout layers B) Using more training data C) Reducing the number of layers D) All of the above

6. What is a convolutional neural network (CNN) primarily used for?

A) Natural language processing tasks B) Image and video recognition tasks C) Time series prediction D) Reinforcement learning

7. What is the vanishing gradient problem?

A) A situation where gradients become too large during training B) A situation where gradients become too small, hindering weight updates C) A problem where data is lost during training D) A situation where the model's accuracy suddenly drops

8. Which of the following is a popular optimization algorithm in deep learning?

A) Stochastic Gradient Descent (SGD) B) Linear Regression C) Principal Component Analysis D) K-Means Clustering

9. What is the function of a loss (or cost) function in a neural network?

A) To measure how well the model is performing B) To update the input data C) To initialize the weights D) To adjust the learning rate

10. What is "dropout" in the context of neural networks?

A) A method for increasing the size of the dataset B) A regularization technique where randomly selected neurons are ignored during training C) An optimization algorithm D) A type of activation function