Machine Learning Hyperparameters and Models

What is the primary goal when dealing with a large dataset in machine learning?

• To discover patterns and make predictions

• To eliminate all outliers

• To increase the size of the dataset

• To reduce the number of data points

Why is selecting a machine learning model not straightforward?

• Because models are too expensive

• Because there are only a few models to choose from

• Because all models perform the same

• Because there are many model types and architectures

What does the batch size parameter determine in training?

• The number of layers in the model

• The number of data points used at once for training

• The number of epochs

• The type of model used

Which of the following is NOT a training parameter?

• Learning rate

• Batch size

• Number of epochs

• Model architecture

What are model parameters in a neural network?

• The number of epochs

• The weights and biases

• The type of activation function

• The learning rate and batch size

What is the role of hyperparameters in the training process?

• They affect the speed and quality of training

• They are not important for training

• They are learned during training

• They are the same as model parameters

What is the prefix used for parameters that are over and above the training process?

• Hyper

• Ultra

• Mega

• Super

What is hyperparameter tuning compared to?

• Tuning a car engine

• Tuning a guitar

• Tuning a radio

• Tuning a piano

How are optimal hyperparameter values determined?

• By using the same values for all models

• Using a fixed algorithm

• Through trial and error

• By guessing

What is the main challenge in selecting hyperparameters?

• They are too expensive to compute

• There is no efficient algorithm to select them

• They do not affect the model performance

• They are the same for all models