Vitamin 7 _ Gradescope
pdf
keyboard_arrow_up
School
University of California, Berkeley *
*We aren’t endorsed by this school
Course
C102
Subject
Mathematics
Date
Apr 3, 2024
Type
Pages
7
Uploaded by GeneralProtonBeaver21
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
1/7
7/7
Questions Answered
Saved at 10:21 PM
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
2/7
Vitamin 7
Q1 Mid-Semester Survey
3 Points
For this question, just share your honest thoughts with us in this Mid-Semester survey. Please take the time to provide feedback so we can improve the course and help you learn as best we can: thank you!
After filling out the Mid-Semester Survey
, input the passcode you received upon completion of the survey in the textbox below to receive credit for this question:
WBUkbtwZVl
Correct
Save Answer
Last saved on Mar 10 at 10:21 PM
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
3/7
Q2 Random Forests
1 Point
Which of the following statements are true?
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
Q3 Parametrics and Nonparametrics
1 Point
True or false: Non-parametric methods are always harder to interpret than parametric ones.
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
After training a random forest model, all the individual
trees must be identical.
A random forest model is more robust (is less variable)
overall than a single decision tree.
One way to increase the diversity of individual decision
trees within a random forest model is to limit the depth of
the trees.
One way to increase the diversity of individual decision
trees within a random forest model is to train each one on
data with only a subset of the features used.
True
False
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
- Access to all documents
- Unlimited textbook solutions
- 24/7 expert homework help
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
4/7
Q4 Activations
1 Point
Which of the following functions are valid activation layers for a neural network?
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
ReLU (restricted linear unit)
f
(
x
) = 6
x
+ 45
Sigmoid function
arctan
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
5/7
Q5 Backpropagation
3 Points
We have a function . We are evaluating this function at the point . Use the following computation graph to compute the gradient using backpropagation.
Q5.1
1 Point
Finish the forward pass of backpropagation. Which of the following correspond to the missing values α and f?
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
f
(
x
,
x
,
x
) =
1
2
3
(
x
+
1
x
) ⋅
2
x
3
(
x
,
x
,
x
) =
1
2
3
(8, −4, 10)
, α
= 4
f
= 14
, α
= −32
f
= −320
, α
= −32
f
= −22
, α
= 4
f
= 40
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
6/7
Q5.2
1 Point
Compute the following partial derivatives and evaluated at .
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
Q5.3
1 Point
Use the partial derivatives you computed to finish computing the gradients and . We are already given .
Correct
Save Answer
Last saved on Mar 10 at 9:59 PM
Save All Answers
Submit & View Submission
d
α
d
f
d
x
1
d
α
(
x
,
x
,
x
)
1
2
3
, =
d
α
d
f
10
=
d
x
1
d
α
1
, =
d
α
d
f
4
=
d
x
1
d
α
1
, =
d
α
d
f
10
=
d
x
1
d
α
−4
, =
d
α
d
f
4
=
d
x
1
d
α
10
d
x
1
d
f
d
x
3
d
f
=
d
x
2
d
f
10
, =
d
x
1
d
f
10
=
d
x
3
d
f
4
, =
d
x
1
d
f
1
=
d
x
3
d
f
4
, =
d
x
1
d
f
1
=
d
x
3
d
f
1
, =
d
x
1
d
f
4
=
d
x
3
d
f
4
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
- Access to all documents
- Unlimited textbook solutions
- 24/7 expert homework help
3/10/24, 10:21 PM
Submit Vitamin 7 | Gradescope
https://www.gradescope.com/courses/711377/assignments/4211918/submissions/new
7/7