Quiz 14

The output range of sigmoid activation function ??

• [-1,1]

• [0,1]

• nothing of them

The output range of SVM function ??

• [-1,1]

• [0,1]

• nothing of them

The output range of tradtional activation function ??

• [-1,1]

• [0,1]

• nothing of them

The goal of Regularization is reduce ...... error

• generalization

• training

• Validation

Increasing the regularization of a linear regression model will ............... the bias.

• Increase

• Decrease

• not affect

• None of the mentioned

.......... is a method for solving for the regression model parameters analytically.

• Gradient descent

• Normal equation

• Expectation_Maximization

• None of the mentioned

If you would run 10-fold cross validation on a dataset of 100 examples, how many

examples you will need for the training set in each run?

• 10

• 90

• 100

• 1000

In Stochastic Gradient Descent used for optimizing your regression model, the model's

parameters would be updated after ..... training examples.

• one

• m

• all

• any of them

You observe the following while fitting a linear regression to the data; As you

increase the model complexity, the test error increases and the training error

decreases. What do you think is the main reason behind this behavior?

• Overfitting

• Underfitting

• Generalization

• High Bias

Which of the following statement(s) is/are true for batch Gradient Decent (GD)?

1) You update a set of parameters in an iterative manner to minimize the error

function.

2) You have to run through all the samples in your training set for a single update

of a parameter in each iteration.

• Only 1

• Only 2

• 1 and 2

• None of the above