בייס

Discuss, Learn and be Happy דיון בשאלות

• «
• 1
• 2
• 3
• »

15. The figure below shows a likelihood function and a posterior distribution. Both are Gaussian, with σposterior=1.2 and σlikelihood=1.5. Assume that the prior is also Gaussian. Which of the following statements is true?

-

1

+

שאלה #22013

		a) The prior is centered to the left of the likelihood function and is narrower.
done		b) The prior is centered to the left of the likelihood function and is broader.
		c) The prior is centered to the right of the likelihood function and is narrower.
		d) The prior is centered to the right of the likelihood function and is broader.

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

The likelihood function is equal to the noise distribution

-

1

+

שאלה #22014

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

The value of the stimulus s that maximizes posterior probability is the value of the measurement x

-

1

+

שאלה #22015

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

We can estimate the distribution of subject’s responses by multiplying the noise distribution with the prior

-

1

+

שאלה #22016

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

If, over the course of an experiment, a Bayesian observer reports one value of the stimulus estimate more often than another value, it means that the prior probability of the former is higher

-

1

+

שאלה #22017

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

18. An observer combines cues A and B. When B becomes more reliable, the observer’s estimate will

-

1

+

שאלה #22018

		a) shift towards A
		b) shift towards B
		c) stay the same
done		d) Insufficient information

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

In modeling speech recognition, the sound and the image can be assumed to be independent of each other

-

1

+

שאלה #22019

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

Conflicts between two measurements generated by a single source do not normally occur in the natural world

-

1

+

שאלה #22020

		True
done		False

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

22. Camouflage works by?

-

1

+

שאלה #22021

done		a. flattening likelihood functions
		b. sharpening posteriors
		c. flattening priors

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

25. Do each of the following represent variability (many trials) or uncertainty (single trial)?

-

1

+

שאלה #22022

done		Noise distribution V, posterior distribution U, estimate distribution V
		Noise distribution U, posterior distribution U, estimate distribution V
		Noise distribution V, posterior distribution U, estimate distribution U
		Noise distribution U, posterior distribution V, estimate distribution V

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

• «
• 1
• 2
• 3
• »