Exam Questions Flashcards by M d Heijer | Brainscape

Q

What is the difference between Gibbs sampling and Metropolis-Hastings random walk?

A

Gibbs sampling is used for when you know the exact distribution, Metropolis-Hastings (random walk) is used for when you just have the kernel.

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

What is the intuition of the formula of the Metropolis-Hastings Independence Chain method?

A

It corrects for the difference between Q(θ) and P(θ)

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

What is the definition of the Posterior Odds Ratio?

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

What is the definition of the Prior odds ratio?

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

What is the definition of the Bayes factor?

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

How is the posterior odds ratio related to the prior odds ratio and the Bayes factor? Give an intuitive explanation.

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

How to calculate the Bayes factor? How to calculate the probabilities for the models?

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly

Q

What is the Jefferys-Lindley-Bartlett paradox?

A

How well did you know this?

1

Not at all

2

3

4

5

Perfectly