Logistic Regression Flashcards

1
Q

Concept

A

Aim is to predict how likely it is that an event will occur that OV = 1 or how likely it is that OV = 0

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

Why use logistic regression and not linear regression?

A

Many assumptions such as linear assumption will be violated

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

Odds

A

Long run ratio of an event happening to an event not happening
Odds (wins) = n (wins) / n (losses) = p (win) / p (lose)

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

What is chi square used for

A

To test significance of the model (bc categorical OV)

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

3 values of model testing

A

Deviance (-2LL)
Pseudo R squared
Hit rate

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

-2LL (log likelihood)

A

How much unexplained information after model has been fitted
(Big values, bad models)

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

Pseudo R squared

A

Explanatory power of the model with PVs compared to null model without PVs
(Higher pseudo R squared the better)

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

Hypotheses chi square

A

H0: newly added PVs, compared to null, have no difference on OV
- beta 1 = beta 2 = … = beta k = 0

H1: at least one beta does not equal 0

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

Hit rate

A

Number of people in dataset we correctly predict the event to occur

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Effect of high cutoff rate

A

LESS 1s MORE 0s
Prediction of extra 0s are incorrect so correct estimates for 0s decrease and increase for 1s

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Effect of low cutoff rate

A

MORE 1s LESS 0s
Prediction of extra 1s are not correct so probability of correct estimates for 1s decrease and increase for 0s

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Exp(B) concept

A

When we increase PV by 1 unit, the odds of that event happening will change by exp(B) factor
Less than one: negative effect
More than one: positive effect
One: no effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Reporting chi square statistic

A

If p is less than 0.05 we reject H0 and conclude that at least one of the betas is not equal to 0, and at least one of the PVs is significant and has an effect on OV, so our model as a whole adds explanatory power compared to the null model

Reject? Equal to 0? Significant? Effect on OV? Adds explanatory power?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Pseudo R squared interpretation

A

Explanatory power of the model with PVs compared to model without PVs
Gives an idea about “how well the PVs in the model fits into the data”
Higher they are more improvement

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Interpreting beta

A

Only see if the effect is positive or negative

Values above 0 have positive effect
Below 0 have negative effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Interpreting exp(B)

A

Exp(B) = odds ratio!
Interprets effect size
When we increase PV by 1 unit, the odds of that event happening (y=1) will change by factor exp (B)