lOMoARcPSD|9933036

Final Lab Exam Attempt review

Board Certified Environmental Scientist Air Resources Certification Exam (AMA

Computer University)

Scan to open on Studocu

Studocu is not sponsored or endorsed by any college or university

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)
 lOMoARcPSD|9933036

Home / My courses / UGRD-CYBS6101-2323T / FINAL EXAMINATION / Final Lab Exam

Started on Thursday, 4 April 2024, 8:51 PM

State Finished
Completed on Thursday, 4 April 2024, 9:06 PM
Time taken 15 mins 10 secs
Marks 39.00/50.00
Grade 78.00 out of 100.00

Question 1

Correct

Mark 1.00 out of 1.00

What is the process of adding additional layers to a backpropagation network called?

Select one:
a. Tuning

b. Optimizing

c. Deepening
d. Regularizing

Question 2

Correct

Mark 1.00 out of 1.00

What is an action in a Markov decision process (MDP)?

Select one:
a. A possible outcome of an action

b. A representation of the future situation

c. A representation of the current situation

d. A representation of the past situation

Question 3

Correct

Mark 1.00 out of 1.00

What is the gradient descent in the backpropagation algorithm?

Select one:
a. The process of maximizing the error between the predicted output and the actual output

b. The process of minimizing the error between the predicted output and the actual output
c. The process of adjusting the weights and biases in the backward direction

d. The process of adjusting the weights and biases in the forward direction

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 4

Incorrect

Mark 0.00 out of 1.00

What is the process of evaluating the performance of a backpropagation network called?

Select one:
a. Validating 

b. Testing
c. Inferring

d. Training

Question 5

Correct

Mark 1.00 out of 1.00

What is an agent in reinforcement learning?

Select one:
a. The labeled data used to train the model

b. The environment in which the learning takes place

c. The predicted outcome of an event

d. The entity that receives rewards or punishments and learns from them

Question 6

Incorrect

Mark 0.00 out of 1.00

What is the main difference between the Quickprop algorithm and the Resilient Propagation (Rprop) algorithm?

Select one:
a. The Quickprop algorithm is more efficient, while the Rprop algorithm is less efficient
b. The Quickprop algorithm uses a variable learning rate, while the Rprop algorithm uses a fixed learning rate 

c. The Quickprop algorithm is less efficient, while the Rprop algorithm is more efficient

d. The Quickprop algorithm uses a fixed learning rate, while the Rprop algorithm uses a variable learning rate

Question 7

Correct

Mark 1.00 out of 1.00

What is a state in reinforcement learning?

Select one:
a. The current condition or situation of the agent
b. The entity that receives rewards or punishments and learns from them

c. The predicted outcome of an event

d. The environment in which the learning takes place

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 8

Correct

Mark 1.00 out of 1.00

What is an output alphabet?

Select one:
a. The set of states that a state machine can be in

b. The set of outputs that a state machine can produce

c. The set of inputs that a state machine can receive

d. The set of transitions that a state machine can make

Question 9

Correct

Mark 1.00 out of 1.00

What is a backpropagation network?

Select one:
a. A type of artificial neural network that uses reinforcement learning

b. A type of artificial neural network that uses unsupervised learning

c. A type of artificial neural network that uses deep learning

d. A type of artificial neural network that uses supervised learning

Question 10

Correct

Mark 1.00 out of 1.00

What is the process of applying an activation function to the dot product of the input layer and the weights called?

Select one:
a. Backpropagation
b. None of the above

c. Activation

d. Forward propagation

Question 11

Correct

Mark 1.00 out of 1.00

What is the main advantage of the RProp algorithm compared to other optimization algorithms?

Select one:
a. It is faster and more efficient
b. It is easier to implement

c. It is more accurate and precise

d. It is more resilient to noise and outliers in the data

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 12

Correct

Mark 1.00 out of 1.00

What is a state in a Markov decision process (MDP)?

Select one:
a. A representation of the current situation

b. A possible outcome of an action

c. A representation of the future situation

d. A representation of the past situation

Question 13

Correct

Mark 1.00 out of 1.00

What is the RProp algorithm's learning rate update rule?

Select one:
a. The learning rate is updated based on the difference between the current and previous iteration's Hessian matrix

b. The learning rate is updated based on the difference between the current and previous iteration's error
c. The learning rate is updated based on the difference between the current and previous iteration's gradient

d. The learning rate is updated based on the difference between the current and previous iteration's weights

Question 14

Correct

Mark 1.00 out of 1.00

What are the weights in the backpropagation algorithm?

Select one:
a. The values that are adjusted in the backward pass
b. The values that are used to determine the output of the neural network

c. The values that are adjusted in the forward pass

d. The values that are used to determine the activation function of the neural network

Question 15

Incorrect

Mark 0.00 out of 1.00

What is the main disadvantage of the Quickprop algorithm?

Select one:
a. It is less efficient than other algorithms
b. It is more sensitive to initialization

c. It requires more computational resources

d. It is more complex than other algorithms 

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 16

Incorrect

Mark 0.00 out of 1.00

What is the main difference between the Quickprop algorithm and the Scaled Conjugate Gradient (SCG) algorithm?

Select one:
a. The Quickprop algorithm is less efficient, while the SCG algorithm is more efficient

b. The Quickprop algorithm is more efficient, while the SCG algorithm is less efficient
c. The Quickprop algorithm uses a variable learning rate, while the SCG algorithm uses a fixed learning rate 

d. The Quickprop algorithm uses a fixed learning rate, while the SCG algorithm uses a variable learning rate

Question 17

Correct

Mark 1.00 out of 1.00

What is a state machine?

Select one:
a. A machine that is in a constant state

b. A machine that changes its state based on a state transition function

c. A machine that can only be in one of two states

d. A machine that can be in an infinite number of states

Question 18

Correct

Mark 1.00 out of 1.00

How does the RProp algorithm adjust the learning rate?

Select one:
a. It increases the learning rate if the error increases and decreases the learning rate if the error decreases
b. It increases the learning rate if the error decreases and decreases the learning rate if the error increases

c. It uses a predetermined set of learning rates for each iteration

d. It uses a fixed learning rate regardless of the error

Question 19

Correct

Mark 1.00 out of 1.00

How is a state transition function typically represented?

Select one:
a. As a table or matrix
b. As a function of state and input

c. As a graph or network
d. As a set of differential equations

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 20

Correct

Mark 1.00 out of 1.00

What is the hidden layer in the backpropagation algorithm?

Select one:
a. The layer that is not visible to the user and processes the input data

b. The layer that produces the final output of the neural network
c. The layer that is visible to the user and processes the input data

d. The layer that determines the activation function of the neural network

Question 21

Correct

Mark 1.00 out of 1.00

How does the Quickprop algorithm handle weight updates that are too large?

Select one:
a. It discards the weight updates

b. It reduces the learning rate

c. It reduces the weight updates

d. It increases the learning rate

Question 22

Correct

Mark 1.00 out of 1.00

What is a disadvantage of using a Q-learning algorithm?

Select one:
a. It requires a large amount of data
b. It cannot handle high-dimensional data

c. It can learn from its mistakes and improve over time

d. It cannot handle continuous state spaces

Question 23

Correct

Mark 1.00 out of 1.00

What is the process of adjusting the weights and biases based on the error in the output layer called?

Select one:
a. Activation
b. None of the above

c. Backpropagation
d. Forward propagation

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 24

Incorrect

Mark 0.00 out of 1.00

What is the main difference between the Quickprop algorithm and the Levenberg-Marquardt (LM) algorithm?

Select one:
a. The Quickprop algorithm is less efficient, while the LM algorithm is more efficient

b. The Quickprop algorithm uses a fixed learning rate, while the LM algorithm uses a variable learning rate
c. The Quickprop algorithm is more efficient, while the LM algorithm is less efficient

d. The Quickprop algorithm uses a variable learning rate, while the LM algorithm uses a fixed learning rate 

Question 25

Correct

Mark 1.00 out of 1.00

What is the RProp algorithm's weight update rule?

Select one:
a. The weight update is inversely proportional to the second derivative of the error function

b. The weight update is inversely proportional to the derivative of the error function
c. The weight update is proportional to the second derivative of the error function

d. The weight update is proportional to the derivative of the error function

Question 26

Correct

Mark 1.00 out of 1.00

What is the RProp algorithm used for?

Select one:
a. Optimization of neural networks
b. Clustering

c. Classification

d. Regression analysis

Question 27

Correct

Mark 1.00 out of 1.00

What is a linear regression algorithm?

Select one:
a. A tree-like model used to make predictions in supervised learning
b. A neural network used in deep learning

c. A clustering algorithm used in unsupervised learning

d. A model used to make predictions in supervised learning by fitting a linear equation to the data

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 28

Incorrect

Mark 0.00 out of 1.00

What is the process of evaluating a feedforward neural network on a set of test data called?

Select one:
a. Testing

b. Validation 
c. Training

d. All of the above

Question 29

Correct

Mark 1.00 out of 1.00

How does a backpropagation network learn?

Select one:
a. By adjusting the weights and biases of the network based on the input and output data

b. By adjusting the architecture of the network based on the input and output data
c. By adjusting the number of neurons in the network based on the input and output data

d. By adjusting the activation function of the network based on the input and output data

Question 30

Correct

Mark 1.00 out of 1.00

What is the main characteristic of the RProp algorithm?

Select one:
a. It uses a stochastic gradient descent
b. It uses a predetermined set of weights

c. It uses a dynamic learning rate

d. It uses a fixed learning rate

Question 31

Incorrect

Mark 0.00 out of 1.00

What is the process of adjusting the weights and biases of a backpropagation network called?

Select one:
a. Validating
b. Optimizing 

c. Testing
d. Training

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 32

Correct

Mark 1.00 out of 1.00

What is the input layer of a feedforward neural network called?

Select one:
a. None of the above

b. Hidden layer
c. Output layer

d. Input layer

Question 33

Incorrect

Mark 0.00 out of 1.00

What is the weight update rule in backpropagation?

Select one:
a. W(i, j) = W(i, j) - alpha * delta(i) * output(j)

b. W(i, j) = W(i, j) * alpha * delta(i) * output(j)

c. W(i, j) = W(i, j) / alpha * delta(i) * output(j) 

d. W(i, j) = W(i, j) + alpha * delta(i) * output(j)

Question 34

Incorrect

Mark 0.00 out of 1.00

What is the main difference between the Quickprop algorithm and the Conjugate Gradient (CG) algorithm?

Select one:
a. The Quickprop algorithm uses a variable learning rate, while the CG algorithm uses a fixed learning rate 
b. The Quickprop algorithm is more efficient, while the CG algorithm is less efficient

c. The Quickprop algorithm is less efficient, while the CG algorithm is more efficient

d. The Quickprop algorithm uses a fixed learning rate, while the CG algorithm uses a variable learning rate

Question 35

Correct

Mark 1.00 out of 1.00

What is the input layer in the backpropagation algorithm?

Select one:
a. The layer that receives the input data and passes it to the hidden layer
b. The layer that produces the final output of the neural network

c. The layer that is not visible to the user and processes the input data
d. The layer that determines the activation function of the neural network

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 36

Correct

Mark 1.00 out of 1.00

What is reinforcement learning?

Select one:
a. A type of deep learning where the model is trained using a large dataset

b. A type of supervised learning where the model is trained using labeled data
c. A type of machine learning where an agent learns through interacting with its environment and receiving rewards or 
punishments
d. A type of unsupervised learning where the model is trained using unlabeled data

Question 37

Correct

Mark 1.00 out of 1.00

What is the process of verifying that a backpropagation network has not overfitted to the training data called?

Select one:
a. Inferring

b. Training
c. Validating

d. Testing

Question 38

Correct

Mark 1.00 out of 1.00

What is the goal of a Markov decision process (MDP)?

Select one:
a. To minimize the expected cumulative cost over a given time horizon
b. To maximize the expected cumulative reward over a given time horizon

c. To minimize the expected cumulative reward over a given time horizon

d. To maximize the expected cumulative cost over a given time horizon

Question 39

Correct

Mark 1.00 out of 1.00

What is the learning rate in backpropagation?

Select one:
a. The process of adjusting the weights and biases in the backward direction
b. The hyperparameter that determines the size of the weight and bias updates

c. The process of adjusting the weights and biases in the forward direction
d. The process of minimizing the error between the predicted output and the actual output

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 40

Correct

Mark 1.00 out of 1.00

What is the goal of the backpropagation algorithm in each iteration?

Select one:
a. To maximize the error between the predicted output and the actual output

b. To minimize the error between the predicted output and the actual output
c. To maximize the error between the predicted output and the actual output in each iteration

d. To minimize the error between the predicted output and the actual output in each iteration

Question 41

Incorrect

Mark 0.00 out of 1.00

What is the process of using a backpropagation network to make predictions called?

Select one:
a. Testing 

b. Training
c. Inferring

d. Validating

Question 42

Correct

Mark 1.00 out of 1.00

How does the Quickprop algorithm improve upon traditional gradient descent algorithms?

Select one:
a. It uses a variable learning rate
b. It uses a smaller learning rate

c. It uses a fixed learning rate

d. It uses a larger learning rate

Question 43

Correct

Mark 1.00 out of 1.00

What is a policy in reinforcement learning?

Select one:
a. The environment in which the learning takes place
b. The strategy or behavior followed by the agent in order to maximize its reward

c. The current condition or situation of the agent

d. The entity that receives rewards or punishments and learns from them

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 44

Correct

Mark 1.00 out of 1.00

How does the RProp algorithm handle the initialization of weights in the optimization process?

Select one:
a. It uses predetermined weights

b. It uses the previous iteration's weights

c. It uses random weights

d. It uses the weights of the best performing iteration

Question 45

Correct

Mark 1.00 out of 1.00

What is a decision tree?

Select one:
a. A clustering algorithm used in unsupervised learning

b. A tree-like model used to make predictions in supervised learning

c. A neural network used in deep learning

d. A linear model used to make predictions in supervised learning

Question 46

Correct

Mark 1.00 out of 1.00

What is the bias update rule in the backpropagation algorithm?

Select one:
a. The process of adjusting the weights and biases in the forward direction
b. The mathematical formula that is used to update the biases based on the gradient descent

c. The process of adjusting the weights and biases in the backward direction

d. The process of minimizing the error between the predicted output and the actual output

Question 47

Correct

Mark 1.00 out of 1.00

What is a reward in reinforcement learning?

Select one:
a. The entity that receives rewards or punishments and learns from them
b. The current condition or situation of the agent

c. A positive or negative feedback received by the agent for its actions

d. The environment in which the learning takes place

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)

 lOMoARcPSD|9933036

Question 48

Incorrect

Mark 0.00 out of 1.00

What is the process of finding the optimal policy in a Markov decision process (MDP) called?

Select one:
a. Policy optimization

b. Bellman optimization
c. Dynamic programming

d. Markov optimization 

Question 49

Correct

Mark 1.00 out of 1.00

What is an example of a reinforcement learning problem?

Select one:
a. Grouping customers into different segments based on their spending habits

b. Predicting the stock price for the next day based on historical data
c. Training a robot to navigate through a maze

d. Determining whether an email is spam or not

Question 50

Correct

Mark 1.00 out of 1.00

What is a state transition function?

Select one:
a. A function that determines the current state of a system based on its next state and output
b. A function that determines the current state of a system based on its previous state and input

c. A function that determines the next state of a system based on its previous state and output

d. A function that determines the next state of a system based on its current state and input

← Final Exam

Jump to...

Downloaded by Aaron Dasma (aaron.dasma666@gmail.com)