Homework answers / question archive /  HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope https://www

 HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope https://www

Computer Science

Share With

 HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 1/10
0/21 Questions Answered
HW6: Probabilistic Reasoning
(Programming Assignment)
Q1 Particle Filtering
75 Points
Q1.1 Task 1.1
25 Points
Nothing to submit here. Just include your code in the zip file you
submit in Q4. Worth 25 points.
Save Answer
Q1.2 Task 1.2
10 Points
Upload your plots here. Don't cherry-pick. We may run your code to
generate plots and compare. Just upload an image of one or more
plots representing typical performance.
Please select file(s) Select file(s)
Particle count
Enter your answer here
Save Answer
Q1.3 Task 1.3: Accuracy vs noise
4CPhooionstesFiles No file chosen
?

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 2/10
Select all true statements.
Save Answer
Q1.4 Task 1.3: Accuracy vs number of particles
4 Points
Select all true statements.
Save Answer
Choose Files No file chosen
Accuracy and noise are directly proportional.
Accuracy and noise are inversely proportional.
Accuracy plateaus as noise increases.
Accuracy plateaus as noise decreases.
Particle filter is accurate with low nonzero noise.
Particle filter is not accurate with low nonzero noise.
There is little to no effect of noise on accuracy.
Accuracy and number of particles are directly proportional.
Accuracy and number of particles are inversely proportional.
Accuracy plateaus as number of particles increases.
Accuracy plateaus as number of particles decreases.
A high number of particles can compensate for a decrease in
accuracy due to noise.
There is little to no effect of number of particles on accuracy.

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 3/10
Q1.5 Task 1.3: Accuracy vs max sensor range
4 Points
Select all true statements.
Save Answer
Q1.6 Task 1.3: Efficiency vs noise
4 Points
Select all true statements.
Save Answer
Choose Files No file chosen
Accuracy and max sensor range are directly proportional.
Accuracy and max sensor range are inversely proportional.
Accuracy plateaus as max sensor range increases.
Accuracy plateaus as max sensor range decreases.
There is little to no effect of max sensor range on accuracy.
Efficiency and noise are directly proportional.
Efficiency and noise are inversely proportional.
Efficiency plateaus as noise increases.
Efficiency plateaus as noise decreases.
Particle filter is efficient with low nonzero noise.
Particle filter is not efficient with low nonzero noise.
There is little to no effect of noise on efficiency.

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 4/10
Q1.7 Task 1.3: Efficiency vs number of particles
4 Points
Select all true statements.
Save Answer
Q1.8 Task 1.3: Efficiency vs max sensor range
4 Points
Select all true statements.
Save Answer
Q1.9 Task 1.4
4 Points
Choose Files No file chosen
Efficiency and number of particles are directly proportional.
Efficiency and number of particles are inversely proportional.
Efficiency plateaus as number of particles increases.
Efficiency plateaus as number of particles decreases.
There is little to no effect of number of particles on efficiency.
Efficiency and max sensor range are directly proportional.
Efficiency and max sensor range are inversely proportional.
Efficiency plateaus as max sensor range increases.
Efficiency plateaus as max sensor range decreases.
There is little to no effect of max sensor range on efficiency.

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 5/10
What probabilistic assumptions are we making about the state
transition model? Select all that are true.
Save Answer
Q1.10 Task 1.4
4 Points
How well do these assumptions match the true state transition model?
Save Answer
Choose Files No file chosen
The state transition model is Markovian.
The state transition model is non-Markovian.
The distribution of next state conditioned on previous state is
uniform.
The distribution of next state conditioned on previous state is
Gaussian.
The Markovian assumption is held.
The Markovian assumption is not held.
The non-Markovian assumption is held.
The non-Markovian assumption is not held.
The uniform assumption is held.
The uniform assumption is not held.
The Gaussian assumption is held.
The Gaussian assumption is not held.

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 6/10
Q1.11 Task 1.4
4 Points
In what way and to what extent is performance affected by making
these assumptions?
Save Answer
Q1.12 Task 1.5
4 Points
Describe some cases in which the particle filter fails to localize the car
in this map, and explain why this happens. Your response should be
less than 5 sentences.
Enter your answer here
Save Answer
Choose Files No file chosen
The Markovian assumption has a positive impact on accuracy.
The Markovian assumption has a negative impact on accuracy.
The non-Markovian assumption has a positive impact on
accuracy.
The non-Markovian assumption has a negative impact on
accuracy.
The uniform assumption has a positive impact on accuracy.
The uniform assumption has a negative impact on accuracy.
The Gaussian assumption has a positive impact on accuracy.
The Gaussian assumption has a negative impact on accuracy.

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 7/10
Q2 Bayesian Networks
25 Points
Q2.1 Task 2.1
5 Points
Drawing of the Bayes net structure:
Please select file(s) Select file(s)
Save Answer
Q2.2 Task 2.2
10 Points
Nothing to submit here. Just include your code in the zip file you
submit in Q4. Worth 10 points.
Save Answer
Q2.3 Task 2.3
5 Points
Optimal overtake condition found by your code:
Save Answer
Q2.4 Task 2.3
5 Points
Mathematical formula:
Choose Files No file chosen
?
? MuchFaster = True, Early = True
? MuchFaster = True, Early = False
? MuchFaster = False, Early = True
? MuchFaster = False, Early = False

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 8/10
Please select file(s) Select file(s)
Save Answer
Q3 Kalman Filtering (Extra Credit)
0 Points
Q3.1 Task 3.1
0 Points
(+5 points) Nothing to submit here. Just include your code in the zip
file you submit in Q4.
Save Answer
Q3.2 Task 3.2
0 Points
(+5 points) Explain how you set your parameters. Your response must
be no more than 10 sentences.
Enter your answer here
Save Answer
Q3.3 Task 3.3
0 Points
(+5 points)
Plot for KF with GPS noise dist = gaussian, GPS gaussian noise var =
10.0
Please select file(s) Select file(s)
Choose Files No file chosen
?
?

4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 9/10
Plot for KF with GPS noise dist = uniform, GPS uniform noise width =
20.0
Please select file(s) Select file(s)
Accuracy vs measurement noise var/width (no more than 3 sentences)
Enter your answer here
Accuracy vs measurement noise distribution (gaussian or uniform) (no
more than 3 sentences)
Enter your answer here
Optional: if you want to reference plots or other illustrations in your
responses above, upload them here:
Please select file(s) Select file(s)
Save Answer
Q3.4 Task 3.4
0 Points
(+5 points) Your responses must be no more than 3 sentences each.
What assumptions does the Kalman filter make about the GPS
measurements?
Enter your answer here
How well do these assumptions match the process that generates the
data?
? ?
4/20/22, 9:41 AM Submit HW6: Probabilistic Reasoning (Programming Assignment) | Gradescope
https://www.gradescope.com/courses/361635/assignments/2004636/submissions/new 10/10
Enter your answer here
In what way and to what extent is performance affected by these
making these assumptions?
Enter your answer here
Optional: if you want to reference plots or other illustrations in your
responses above, upload them here:
Please select file(s) Select file(s)
Save Answer
Q4 Code submission
0 Points
Submit your zip file here as described in the project PDF:
https://github.com/jdkanu/cmsc421-
p3/blob/main/Programming_Assignment_3__Probabilistic_Reasoning.pdf
Please select file(s) Select file(s)
Save Answer
Save All Answers Submit & View Submission ?
? ?