CS-7638 Final Actual Exam Newest 2026-2027
With Complete 100 Questions And Correct
Detailed Answers| Brand New Version!
Is it possible for particles to never get sampled, even if they have a high
weight? - ANSWER-Yes
Can particle filters be used in continuous spaces? - ANSWER-Yes, they
are usually used for continuous spaces
The weight associated with a particle is calculated from: - ANSWER-The
probability of the observed measurements given the state of the
particle
If a single noisy measurement throws off your particle filter, what can
you change in your particle filter tuning parameters to potentially
improve the situation? - ANSWER-1. Increase the number of particles
2. Increase sigma of your importance weighting function
What is the probability of never resampling a particle? - ANSWER-(1 -
Probability of Resampling that Particle) ^ (Number of Independent
Particles Being Drawn)
1|Page
, Will orientation or heading never play a role in Particle Filters? -
ANSWER-Orientation will eventually matter
True or False: A particle filter with no motion noise, no fuzzing, and a
constant number of resampled particles will usually have all particles
converge to a single point - ANSWER-No motion noise and no fuzzing
will cause particles to be resampled from existing locations, leading to
2+ particles initialized at the same exact spot. As time goes on, the less-
probable locations will fall off, and no new locations will be added until
all particles are on top of each other at the same point. This is why
fuzzing was necessary in the particle filter project -- to add exploration.
Does the re-sampling wheel work when sample size is not equal to
population size? - ANSWER-Yes
Is the particle with the highest weight always sampled? - ANSWER-No
Can the same particle be picked twice in a row using the resampling
wheel? - ANSWER-Yes
What is the error equation? - ANSWER-y = Z - Hx
This equation represents the error by calculating the difference
between the latest measurement and the predicted location
2|Page
With Complete 100 Questions And Correct
Detailed Answers| Brand New Version!
Is it possible for particles to never get sampled, even if they have a high
weight? - ANSWER-Yes
Can particle filters be used in continuous spaces? - ANSWER-Yes, they
are usually used for continuous spaces
The weight associated with a particle is calculated from: - ANSWER-The
probability of the observed measurements given the state of the
particle
If a single noisy measurement throws off your particle filter, what can
you change in your particle filter tuning parameters to potentially
improve the situation? - ANSWER-1. Increase the number of particles
2. Increase sigma of your importance weighting function
What is the probability of never resampling a particle? - ANSWER-(1 -
Probability of Resampling that Particle) ^ (Number of Independent
Particles Being Drawn)
1|Page
, Will orientation or heading never play a role in Particle Filters? -
ANSWER-Orientation will eventually matter
True or False: A particle filter with no motion noise, no fuzzing, and a
constant number of resampled particles will usually have all particles
converge to a single point - ANSWER-No motion noise and no fuzzing
will cause particles to be resampled from existing locations, leading to
2+ particles initialized at the same exact spot. As time goes on, the less-
probable locations will fall off, and no new locations will be added until
all particles are on top of each other at the same point. This is why
fuzzing was necessary in the particle filter project -- to add exploration.
Does the re-sampling wheel work when sample size is not equal to
population size? - ANSWER-Yes
Is the particle with the highest weight always sampled? - ANSWER-No
Can the same particle be picked twice in a row using the resampling
wheel? - ANSWER-Yes
What is the error equation? - ANSWER-y = Z - Hx
This equation represents the error by calculating the difference
between the latest measurement and the predicted location
2|Page
