Weekly Quiz 2 Boosting Ensemble Techniques and Model Tuning Great Learning PDF

Go Back to Ensemble Techniques and Model Tuning
Course Content
Weekly Quiz 2 - Boosting

Type : Graded Quiz
Attempts : 1/1
Questions : 10
Time : 45m
Due Date : CDT
Your Marks : 10/10
Instructions
Attempt History
Attempt #1 Marks: 10
Q No: 1 Correct Answer
Marks: 1/1
Which of the following statement(s) is/are true for bagging and boosting?
A) Bagging: Weak learners are built in parallel

Boosting: Weak learners are built in sequence one after the other
B) Bagging: More weight to those weak learners with better performance

Boosting: Each weak learner has equal weight in the final prediction.
C) Bagging: Samples are drawn from the original dataset with replacement to train each
individual weak learner
This study source was downloaded by 100000854545503 from CourseHero.com on 04-06-2024 15:43:41 GMT -05:00
https://www.coursehero.com/file/209580356/Weekly-Quiz-2-Boosting-Ensemble-Techniques-and-Model-Tuning-Great-Learningpdf/
Boosting: Subsequent samples have more weight of those observations which had relatively
higher errors in previous weak learners
D) Bagging: Random forest is a special type of bagging technique

Boosting: Adaboost is a special type of boosting technique
A, B, and C
B, C, and D
A, C, and D You Selected
A and D
In bagging, weak learners are built in parallel, and samples are drawn from the original dataset
with replacement to train each individual weak learner. Example - Random Forest
In boosting, weak learners are built in sequence one after the other, and subsequent samples
have more weight of those observations which had relatively higher errors in previous weak
learners. Example - Adaboost
Marks: 1/1
Which of the following statement(s) is/are true about the Gradient Boosting trees?
1. Gradient Boosting trees work on residuals instead of changing the weights of observations.
2. By fitting new models to the residuals, the overall learner gradually improves in areas where
residuals are initially high.
1 and 2 You Selected
Only 1
Only 2
None
Gradient boosting trees work by fitting new models to the residuals and thereby gradually
improving the overall learner in areas where residuals are initially high.
Marks: 1/1
Alpha is a hyperparameter that is used while changing the weights of the samples in an AdaBoost
model.
Which of the following statement(s) is/are correct?
A) The weight of a sample is decreased if it is incorrectly classified by the previous weak learner.
B) The weight of a sample is increased if it is incorrectly classified by the previous weak learner.
C) The alpha can be both positive or negative
D) The alpha cannot be negative
Only A
A and B
B and C You Selected
Only D
The weight of a sample is increased if it is incorrectly classified by the previous weak learner.
Depending on whether the weight should be increased or decreased, the alpha can be both
positive or negative.
Marks: 1/1
What is the correct sequence of steps for the gradient boosting algorithm?
A. Calculate the residual for each observation.
B. Update all predictions using previous probabilities and new output values
C. Initialize the model with an initial prediction for all observations
D. Repeat steps by creating a new tree until the maximum number of estimators is reached
E. Build a tree and calculate the output value for each leaf node
A -> B -> C - > D -> E
B -> C-> E-> D-> A
C -> A -> B -> D -> E
C -> A -> E -> B -> D You Selected
The correct sequence -
Initialize the model with an initial prediction for all observations.
Calculate the residual for each observation.
Build a tree and calculate the output value for each leaf node.
Update all predictions using previous probabilities and new output values.
Repeat steps by creating a new tree until the maximum number of estimators is reached.
Marks: 1/1
Which of the following algorithms do NOT have a ‘learning rate’ parameter/hyperparameter?
Gradient Boosting
XGBoost
Random Forest You Selected
All of these
Gradient Boosting and XGBoost are similar algorithms and have a ‘learning rate’
parameter/hyperparameter to reduce overfitting.
Marks: 1/1
Which hyperparameter of XGBoost can be used to deal with the imbalance in data?
gamma
learning_rate
scale_pos_weight You Selected
colsample_by_node
In XGBoost, scale_pos_weight controls the balance of positive and negative weights and is
used to deal with imbalanced classes.
Marks: 1/1
Which of the following statements is false about XGBoost?
XGBoost is a boosting technique
XGBoost builds upon the idea of gradient boosting algorithm with some modifications
It provides features that help in better computing like parallelization, cache optimization,
out of core computing and distributed computing
XGBoost cannot handle missing values internally You Selected
XGBoost has the ability to handle missing values internally.
Marks: 1/1
Which of the following statement(s) is/are correct for boosting technique?
A) Weight of observation, to be selected while building the next weak learner, increases if the
observation is correctly classified
B) Weight of observation, to be selected while building the next weak learner, increases, if the
observation is incorrectly classified
C) Weight of observation, to be selected while building the next weak learner, decreases if the
observation is correctly classified
D) Weight of observation, to be selected while building the next weak learner, decreases if the
observation is incorrectly classified
A and B
B and C You Selected
C and D
A and D
Weight of observation, to be selected while building the next weak learner, increases, if the
observation is incorrectly classified, and the weight decreases if the observation is correctly
classified.
Marks: 1/1
Adaboost takes a weighted voting/weighted average of the weak learners for the final prediction.
True You Selected
False
Adaboost takes a weighted voting/weighted average of the weak learners for the final
prediction.
Marks: 1/1
Which of the following is/are true:
1. Bagging and boosting use heterogeneous learners (different algorithms as different weak
learners).
Comments:
+ Add comments
Previous Next
Proprietary content.©Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited.
© 2023 All rights reserved
Privacy Terms of service Help
Powered by TCPDF (www.tcpdf.org)

Weekly Quiz 2 Boosting Ensemble Techniques and Model Tuning Great Learning PDF

Uploaded by

Copyright:

Available Formats

Weekly Quiz 2 Boosting Ensemble Techniques and Model Tuning Great Learning PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Weekly Quiz 2 Boosting Ensemble Techniques and Model Tuning Great Learning PDF

Uploaded by

Copyright:

Available Formats

Go Back to Ensemble Techniques and Model Tuning

Weekly Quiz 2 - Boosting

Your Marks : 10/10

Q No: 1 Correct Answer

A) Bagging: Weak learners are built in parallel

B) Bagging: More weight to those weak learners with better performance

D) Bagging: Random forest is a special type of bagging technique

A, C, and D You Selected

Q No: 2 Correct Answer

Q No: 3 Correct Answer

Which of the following statement(s) is/are correct?

B and C You Selected

Q No: 4 Correct Answer

A -> B -> C - > D -> E

B -> C-> E-> D-> A

C -> A -> B -> D -> E

C -> A -> E -> B -> D You Selected

The correct sequence -

Initialize the model with an initial prediction for all observations.

Calculate the residual for each observation.

Q No: 5 Correct Answer

Random Forest You Selected

Q No: 6 Correct Answer

scale_pos_weight You Selected

Q No: 7 Correct Answer

XGBoost cannot handle missing values internally You Selected

XGBoost has the ability to handle missing values internally.

Q No: 8 Correct Answer

Which of the following statement(s) is/are correct for boosting technique?

B and C You Selected

Q No: 9 Correct Answer

True You Selected

Q No: 10 Correct Answer

Which of the following is/are true:

© 2023 All rights reserved

Privacy Terms of service Help

You might also like