Week 8 Flashcards

1
Q

Wat geeft de foutmaat e aan in het berekenen van E.out?

A

Welke elementen van y op elkaar lijken.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Geeft een grote H meer kans op overfitten?

A

Ja

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Welk algorithme werkt niet door expliciet te zoeken naar een hypothese die een bepaalde score optimaliseert?

A

k-nearest neighbors.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Wat is het verschil tussen hard en soft clustering?

A

Bij hard is er geen overlap, bij soft wel, tussen clusters.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Does clustering label the clusters?

A

no

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

what does k represent in k-means?

A

the number of clusters

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are two internal evaluation metrics?

A

Coherence: average/variance of distance of points to cluster centroid
Separation: average distance of distance of points to other clusters

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Purity

A

External evaluation metric:

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

confusion matrix

A

external evaluation metric (tp.fp.fn.tn)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly