chapter3-sml

What is the key point of a parametric model?

• It is a non-parametric model

• It depends on the training data for predictions

• It contains some parameters that are learned from training data

• It does not involve any parameters

What is regression in supervised learning?

• Learning the relationships between input variables and a categorical output variable

• Learning the relationships between input variables and a numerical output variable

• Learning the relationships between output variables and a numerical input variable

• Learning the relationships between output variables and a categorical input variable

What is the assumption made about the noise term in linear regression?

• It is constant and does not vary

• It is not considered in the model

• It has a mean value of one and is dependent on the input variables

• It has a mean value of zero and is independent of the input variables

What is the purpose of the intercept term in the linear regression model?

• To account for any non-zero mean in the noise term

• To account for random errors in the data not captured by the model

• To make the model more complex

• To reduce the number of parameters in the model

What is the least squares cost function used for in linear regression?

• Maximizing the sum of squared errors between predictions and actual values

• Minimizing the sum of squared errors between predictions and actual values

• Maximizing the sum of absolute errors between predictions and actual values

• Minimizing the sum of absolute errors between predictions and actual values

What does it mean in practice if X^TX is not invertible?

• The matrix X^TX is non-singular

• The matrix X^TX is singular

• The matrix X^TX does not have a unique inverse

• The matrix X^TX has a unique inverse

What is the goal of maximum likelihood solution in linear regression?

• To find the value of theta that makes observing y as likely as possible

• To maximize the sum of squared errors between predictions and actual values

• To find the value of theta that minimizes the likelihood of observing y

• To minimize the sum of squared errors between predictions and actual values

How can categorical input variables be handled in linear regression?

• By creating dummy variables or using one-hot encoding

• By ignoring them completely

• By excluding them from the model

• By converting them into numerical variables

What is the purpose of constructing a model for voting preferences using probabilities?

• To enforce a specific voting behavior in a population group

• To model the uncertainty in voting preferences among individuals in a population group

• To determine the most popular political party in a population group

• To predict the exact voting preferences of each individual in a population group

What does the softmax function do in multiclass logistic regression?

• It ensures that the output vector sums to one and each element is non-negative

• It transforms the input data into a higher-dimensional space

• It enforces a linear decision boundary for classification

• It restricts the model to only predict two classes in a multiclass problem