Question 1

If data are normalized or scaled, which model performance metric will NOT change?

Accepted Answer

A)  MSE 
B)  RMSE 
C)  R-squared 
D)  All of the above will change 
E)  None of the above will change 
A)  MSE 
B)  RMSE 
C)  R-squared 
D)  All of the above will change 
E)  None of the above will change C

Question 2

Which is possible with the "caret" package?

Accepted Answer

A)  Running a nnet model 
B)  Running a neuralnet model 
C)  Both A and B 
D)  Neither A nor B 
A)  Running a nnet model 
B)  Running a neuralnet model 
C)  Both A and B 
D)  Neither A nor B C

Question 3

Explain "gradient descent" and what it does in a neural network algorithm.

Accepted Answer

Gradient descent is the part of the neural network algorithm that adjusts the weights of neurons at each iteration.
More broadly, the gradient descent algorithm estimates how steeply error will be reduced or increased by a given change in a weight, then selects the change corresponding to the estimated steepest decline in error. The amount estimated by gradient descent is the "learning rate". While this process usually works well, gradient descent can lead to locally suboptimal minimization of error. To avoid suboptimal solutions, some software incorporates noise to jog the algorithm out of local minima.

Question 4

The plotnet command is from which R package?

Accepted Answer

A)  nnet 
B)  neuralnet 
C)  caret 
D)  NeuralNetTools 
A)  nnet 
B)  neuralnet 
C)  caret 
D)  NeuralNetTools

Question 5

Which package supports the Olden method of assessing the relative importance of predictors in a neural network model?

Accepted Answer

A)  NeuralNetTools 
B)  neuralnet 
C)  nnet 
D)  caret 
A)  NeuralNetTools 
B)  neuralnet 
C)  nnet 
D)  caret

Question 6

What was a "violin plot" used for in Chapter 7, for the "iris" example?

Accepted Answer

A)  to estimate variable importance 
B)  to obtain graphical insight into how separable the three classes of Species are on the four predictor variables. 
C)  to visualize the results of a neuralnet model 
D)  to visualize the results on a nnet model 
A)  to estimate variable importance 
B)  to obtain graphical insight into how separable the three classes of Species are on the four predictor variables. 
C)  to visualize the results of a neuralnet model 
D)  to visualize the results on a nnet model

Question 7

What is NOT an advantage of running a neural model under caret?

Accepted Answer

A)  optimization of all parameters is built in 
B)  tuning of some parameters is built in 
C)  cross-validation is built in 
D)  All of the above are advantages 
A)  optimization of all parameters is built in 
B)  tuning of some parameters is built in 
C)  cross-validation is built in 
D)  All of the above are advantages

Question 8

Unlike linear regression, neural networks can handle nonlinearities and interaction effects even when these are not explicit input variables.

Accepted Answer

A) True 
 B)False

Question 9

Neural networks may achieve better results if data are normalized or scaled.

Accepted Answer

A) True 
 B)False

Question 10

Neural networks may be represented by an equation or instruction set.

Accepted Answer

A) True 
 B)False

Question 11

Axons, dendrites, and synapses are studied by what discipline, which is the inspiration for artificial neural networks? ___________________.

Accepted Answer

The answer of Axons, dendrites, and synapses are studied by...

Question 12

Cross-validation of a neural network (or other) model requires …

Accepted Answer

A)  One or more test subsets 
B)  One or more hold-out subsets 
C)  One or more validation subsets 
D)  All of the above 
A)  One or more test subsets 
B)  One or more hold-out subsets 
C)  One or more validation subsets 
D)  All of the above

Question 13

What is UNTRUE of hidden layers of neurons in a neural network model?

Accepted Answer

A)  Depending on the software used, there may be more than one hidden layer. 
B)  It is possible to design a neural network with direct connections from the input to the output layer, skipping the hidden layer. 
C)  The more neurons in hidden layers, the better the resulting model fit. 
D)  Activation rules govern the firing of neurons. 
A)  Depending on the software used, there may be more than one hidden layer. 
B)  It is possible to design a neural network with direct connections from the input to the output layer, skipping the hidden layer. 
C)  The more neurons in hidden layers, the better the resulting model fit. 
D)  Activation rules govern the firing of neurons.

Question 14

Which package brings to R some of the functionality of the Python environment, which is often preferred for large-scale neural network analysis?

Accepted Answer

A)  bigdata 
B)  mlr3keras 
C)  caret 
D)  neuralnet 
A)  bigdata 
B)  mlr3keras 
C)  caret 
D)  neuralnet

Question 15

What makes a neural network model "supervised"?

Accepted Answer

A)  input data in the test set includes the correct output values 
B)  input data in the training set includes the correct output values 
C)  input data in the test set includes the correct input values 
D)  input data in the training set includes the correct input values 
A)  input data in the test set includes the correct output values 
B)  input data in the training set includes the correct output values 
C)  input data in the test set includes the correct input values 
D)  input data in the training set includes the correct input values

Question 16

It is impossible to assess the relative importance of input variables in neural networks, which are a "black box".

Accepted Answer

A) True 
 B)False

If data are normalized or scaled, which model performance metric will NOT change?

Which is possible with the "caret" package?

Explain "gradient descent" and what it does in a neural network algorithm.

The plotnet command is from which R package?

Which package supports the Olden method of assessing the relative importance of predictors in a neural network model?

What was a "violin plot" used for in Chapter 7, for the "iris" example?

What is NOT an advantage of running a neural model under caret?

Unlike linear regression, neural networks can handle nonlinearities and interaction effects even when these are not explicit input variables.

Neural networks may achieve better results if data are normalized or scaled.

Neural networks may be represented by an equation or instruction set.

Axons, dendrites, and synapses are studied by what discipline, which is the inspiration for artificial neural networks? ___________________.

Cross-validation of a neural network (or other) model requires …

What is UNTRUE of hidden layers of neurons in a neural network model?

Which package brings to R some of the functionality of the Python environment, which is often preferred for large-scale neural network analysis?

What makes a neural network model "supervised"?

It is impossible to assess the relative importance of input variables in neural networks, which are a "black box".

Using and Abusing Data Analytics in Social Science

Statistical Analytics With R Part 1

Statistical Analytics With R Part 2

Classification and Regression Trees in R

Random Forests

Modeling and Machine Learning

Network Analysis

Text Analytics

Filters

Exam 7: Neural Network Models and Deep Learning

If data are normalized or scaled, which model performance metric will NOT change?

Which is possible with the "caret" package?

Explain "gradient descent" and what it does in a neural network algorithm.

The plotnet command is from which R package?

Which package supports the Olden method of assessing the relative importance of predictors in a neural network model?

What was a "violin plot" used for in Chapter 7, for the "iris" example?

What is NOT an advantage of running a neural model under caret?

Unlike linear regression, neural networks can handle nonlinearities and interaction effects even when these are not explicit input variables.

Neural networks may achieve better results if data are normalized or scaled.

Neural networks may be represented by an equation or instruction set.

Axons, dendrites, and synapses are studied by what discipline, which is the inspiration for artificial neural networks? ___________________.

Cross-validation of a neural network (or other) model requires …

What is UNTRUE of hidden layers of neurons in a neural network model?

Which package brings to R some of the functionality of the Python environment, which is often preferred for large-scale neural network analysis?

What makes a neural network model "supervised"?

It is impossible to assess the relative importance of input variables in neural networks, which are a "black box".

Using and Abusing Data Analytics in Social Science

Statistical Analytics With R Part 1

Statistical Analytics With R Part 2

Classification and Regression Trees in R

Random Forests

Modeling and Machine Learning

Network Analysis

Text Analytics

Filters