Deep Learning Basics Quiz

What is the purpose of the loss function in deep learning?

• To determine the number of hidden layers

• To iteratively search for and update parameter weights

• To calculate the learning rate

• To initialize the neural network

Which loss function is commonly used for regression problems in deep learning?

• ReLU activation function

• Sigmoid activation function

• Cross-entropy cost function

• Quadratic cost function

What does the negative gradient of a multivariable function mean, in the context of deep learning?

• The direction in which the function increases fastest

• The direction of the fastest function growth

• The direction of the slowest function growth

• The direction in which the function decreases fastest

What is the learning rate in the context of the batch gradient descent algorithm?

• The rate at which the input data is processed

• The rate at which the target output is achieved

• The rate at which weights are adjusted according to the error gradient

• The rate at which the loss function is minimized

What is the core approach of the gradient descent method in deep learning?

• To maximize the loss function

• To ignore the loss function

• To minimize the loss function

• To keep the loss function constant

What is the purpose of activation functions in a neural network?

• To transform the neural network into a linear model

• To add linear factors to the neural network

• To add nonlinear factors to the neural network

• To reduce the computation workload

What is the main cause of the vanishing gradient problem in deep neural networks?

• Insufficient data

• The repeated multiplication of small values during backpropagation

• Very big amount of data

• Large number of feature dimensions

Which technique is applied in deep learning to prevent overfitting?

• Optimizer

• Normalization

• Dataset Cleansing

• Inference

What is the purpose of the Softmax function in a neural network?

• To reduce the underfitting problem

• To reduce the computation workload

• The softmax function is often used as the output layer of a multiclass classification task

• To reduce the overfitting problem

What is the solution to overfitting in deep learning that involves randomly discarding some neurons?

• Dropout

• Early stopping

• Normalization

• Data augmentation