Transcript Since all the distributions have n=50, those with more variability are

An Introduction to Classification
•
Classification vs. Prediction
•
Classification & ANOVA
•
Classification Cutoffs, Errors, etc.
•
Multivariate Classification & Linear
Discriminant Function
An Introduction to Classification
Let’s start by reviewing what “prediction” is…
Using a person’s scores on one or more variables to make a
“best guess” of the that person’s score on another
variable (the value of which isn’t known)
Classification is very similar …
Using a person’s scores on one or more variables to make a
“best guess” of the category to which that person belongs
(when the category type isn’t known).
The difference -- a language “convention”
• if the “unknown variable” is quantitative -- its called prediction
• if the “unknown variable” is qualitative -- its called classification
How does classification work???
Let’s start with an “old friend” -- ANOVA
In its usual form…
• There are two qualitatively different IV groups
• naturally occurring or “created” by manipulation
• A quantitative DV
• H0: MeanG1 = Mean G2
• Rejecting H0: tells us
• There is a relationship between the grouping and DV
• Groups represent populations with different means on the DV
• Knowing what group a person in allows us to guess their DV
score -- mean of that group
Let’s review in a little more detail…
Remember the formula for the ANOVA F-test
variation between groups
size of the mean difference
F = ----------------------------------- = --------------------------------------variation within groups
variation within groups
In words -- F compares the mean difference to the variability
around each of those means
Which of the following will produce the larger F-test ?
The two data sets have the same means, mean difference & N
but the difference is….
Data #2
(@ n = 50)
Data #1
(@ n = 50)
group 1 mean = 30
std dev = 5
group 1 mean = 30
std dev = 15
group 2 mean = 50
std dev = 5
group 2 mean = 50
std dev = 15
Graphical depictions of these data show that the size of F relates to
the amount of overlap between the groups
Data #1
0
Larger F = more
consistent grp dif
10
20
30
40
50
70
Smaller F = less
consistent grp dif
Data #2
0
60
10
20
30
40
50
60
70
80
Notice: Since all the distributions have n=50, those with more variability
are not as tall -- all 4 distributions have the same area
Let’s consider that last one “in reverse”…
Could knowing the person’s score help tell us what qualitative
group they are in? …to “classify” them to the proper group?
an Example…
Research has revealed a statistical relationship between the number of
times a person laughs out loud each day (quant variable) and whether
they are depressed or schizophrenic (qual grouping variable).
Mean laughsDepressed = 4.0 Mean laughsSchizophrenic = 7.0
F(1,34) = 7.00, p < .05
A new (as yet undiagnosed) patient laughs 11 times the first day
what’s your “classification” depressed or schizophrenic?
Another patient laughs 1 time -- your “classification”?
A third new patient laughs 5 times -- your “classification”?
Why were the first two “gimmies” and the last one not?
• When the groups have a mean difference, a score beyond one
of the group means is more likely to belong to that group
than to belong to the other group (unless stds are huge)
• someone who laughs more than the mean for the
schizophrenic group is more likely to be schizohrenic
than to be depressed
• someone who laughs less than the mean of the
depressive group is more likely to be depressed
than to be schizophrenic
• Even when the groups have a mean difference, a score between
the group means is harder to correctly classify (unless stds
are miniscule)
• someone with 5-6 laughs are hardest to classify, because
several depressed and schizophrenic folks have this score
Here’s a graphical depiction of the clinical data...
X
18 dep. patients
mean laughs = 4.0
o
x x xo o o
x x x ox ox o o o
18 schiz. patients
mean laughs = 7.0
x x x ox ox ox ox xo ox o o o
laughs -->
0 1 2 3 4 5 6 7 8 9 0 1 2
Looking at this, its easy to see why we would be ...
• confidant in an assignment based on 11 laughs
• no depressed patients had a score that high
• confident in an assignment based on 1 laugh
• no schizophrenic patients had a score that low
• lacking confidence in an assignment based on 5 or 6 laughs
• several depressed & schizophrenic patients had 5 or 6
The process of prediction required two things…
• that there be a linear relationship between the predictor and the
criterion (reject H0: r = 0)
• a formula (y’ = bx + a) to “translate” a predictor score into an
estimate of a criterion variable score
Similarly, the process of classification requires two things …
• a statistical relationship between the predictor (DV) & criterion
(reject H0: M1 = M2)
• a cutoff to “translate” a person’s score on the predictor (DV)
into an assignment to one group or the other
• where should be place the cutoff???
• Wherever gives us the most accurate classification !!
X
18 dep. patients
mean laughs = 4.0
o
x x xo o o
18 schiz. patients
x x x ox ox o o o
mean laughs = 7.0
x x x ox ox ox ox o
x ox o o o
laughs -->
0 1 2 3 4 5 6 7 8 9 0 1 2
1 1 1
When your groups are the same size and your group score
distributions are symmetrical, things are pretty easy…
• place the cutoff at a position equidistant from the group means
• here, the cutoff would be 5.5 -- equidistant between 4.0 and 7.0
• anyone who laughs more than 5.5 times would be “assigned”
as schizophrenic
• anyone who laughs fewer than 5.5 times would be “assigned”
as depressed
o
x
x x xo o o
18 schiz. patients
x x x ox ox o o o
mean laughs = 7.0
18 dep. patients
mean laughs = 4.0
x x x ox ox ox ox o
x ox o o o
laughs -->
0 1 2 3 4 5 6 7 8 9 0 1 2
1 1 1
We can assess the accuracy of the assignments by building a
“reclassification table”
Actual Diagnosis
Assignment
Depressed
Schizophrenic
Depressed Schizophrenic
14
4
4
14
reclassification accuracy would be 28/36 = 77.78%
When considering simple regression/prediction, we wanted to be
able to compare two potential predictors to determine if one would
be better -- we used Steiger’s Z-test of H0: ry,x1= ry,x2
How do we compare two potential classification variables to
determine if one is a better basis for accurate classification ?
• We do it the same way (with one intermediate step)
• As you might remember from ANOVA, we can express the
“effect size” associated with any F as r (or  - same thing)
• r =  [ F / (F + dferror)]
• So, to compare two potential classification variables
• compute the ANOVA for each variable (on same sample)
• convert each F to r
• compare the r values using Steiger’s Z-test
• remember that you’ll need the correlation between the two
classification variables (rx1,x2)
An example,
Which provides better classification between schiz vs. depression,
# times laughing out loud, or score on a “depression scale”?
• For laughing out loud F(1,34) = 7.00 -- translates to
r =  [ F / (F + dferror)] =  [ 7.0 / (7.0 + 34)] = .413
• For the depression scale F(1,34) = 4.0 -- translates to
r =  [ F / (F + dferror)] =  [ 4.0 / (4.0 + 34)] = .324
• # laughs and depression scale scores are correlated r = .35
• So using Steiger’s Z-test
Z = .495, p > .05
• so there is no advantage of using one of these predictors
over the other
• the apparent difference in r is not greater than chance
When considering simple regression/prediction, we wanted to be
able to compare the same correlation in two different populations - we used Fisher’s Z-test of H0: ry,x1(pop1) = ry,x1(pop2)
How compare the same correlations in two populations to
determine if one population would have more accurate
assignments ?
As you might remember from ANOVA, we can express the
“effect size” associated with any F as r (or  - same thing)
• r =  [ F / (F + dferror)]  for each population
So, to compare two potential classification variables
• compute the ANOVA for each sample
• convert each F to r
• compare the r values using Fisher’s Z-test
An example,
Does # times laughing out loud discriminate between schiz vs.
depression better for in-patients or for out-patients.
• For in-patients F(1,48) = 12.00 -- translates to
r =  [ F / (F + dferror)] =  [ 12.0 / (12.0 + 48)] = .447
Fisher’s Z transformation of = .45 is .485
• For the out-patients F(1,88) = 2.0 -- translates to
r =  [ F / (F + dferror)] =  [ 2.0 / (2.0 + 88)] = .149
• Fisher’s Z = 1.86
• Z(p=.05) = 1.96, so retain H0: -- same discriminability in both pops
Getting ready for ldf…
• multiple regression works better than simple regression because
a y’ based on multiple predictors is a better estimate of y
than a y’ based on a single predictor
• similarly, classification based on multiple predictors will do better
than classification based on a single predictor
• but, how to incorporate multiple predictors into a classification ??
• Like with multiple regression, multiple variables (Xs) are each
given a weighting and a constant is added
• ldf = b1* X1 + b2* X2 + b3* X3 + a
• the composite variable is called a linear discriminant function
• function -- constructed from another variables
• linear -- linear combination of linearly weighted vars
• discriminant -- weights are chosen so that the resulting
has the maximum possible F-test between the groups
So, how does this all work ???
• We start with a grouping variable and a set of quantitative (or
binary) predictors (what would be DVs if doing ANOVAs)
• using an algorithm much like multiple regression, the bivariate
relationship of predictor to the grouping variable & the
collinearities among the predictors are all taken into
account and the weights for the ldf formula are derived
• remember this ldf will have the largest possible F value
between the groups
• a cutoff value for the ldf is chosen the cutoff is chosen (more
fancy computation) to maximize % correct reclassification
• to “use” the formula
• a person’s values on the variables are put into the formula
& their ldf score is computed
• their score is compared to the cutoff, and they are assigned
to one group or the other
“ldf Questions” (very parallel to “regression questions”)
• ANOVA -- can groups be discriminated using the quant var?
• Does one quant variable work better than another to
discriminate the groups? (Steiger’s Z-test to compare
the r for the two quant variables
• Does a quant variable better discriminate between groups for
one population than another? (Fisher’s Z-test to
compare the r from the two populations)
• ldf - can groups be discriminated using a combination of vars?
• Comparison of nested ldf models (X2-change test & McNemar’s X2)
• Comparison of non-nested ldf models (McNemar’s X2)
• Comparison of models across populations (R2 & McNemar’s X2)
• Comparison of models across classification rules (McNemar’s X2)