ROC curves can help us determine the effectiveness of predictive models. This presentation gives an introduction of how to understand and use ROC curves to maximize impact.
2. What is ROC?
Receiver Operating Characteristic
Systematically
trade off detection against false alarm Using
You woke me
up at 3 am!!!
Wake up,
you’re late for
class!!!
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
3. A Brief History of ROC Curves
• Developed by electrical engineers and radar
operators during WWII to detect enemy airplanes
vs. geese.
• Illustrates the performance of binary classifiers -
elements in a set divided into two groups
• Compares trade-offs between detection and false
alarm rate
• Now used in many fields
• Psychology
• Medicine and biometrics
• More recently in machine learning and data mining
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
4. Detection vs. False Alarm
• Detec7on/sensi7vity/true
posi7ve
rate
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
measures
how
many
true
posi0ve
cases
are
correctly
detected
• False
alarm/specificity/false
posi7ve
rate
measures
the
number
of
false
alarms
• Tradeoff:
Usually
can
op0mize
for
one
but
not
both
• Example:
Disease
detec0on
• Sacrifice
false
alarm
for
detec0on
if
cost
of
missed
detec0on
is
alarmingly
high
5. How is ROC Generated?
Features à Scores à PDF à ROC
Model
GPA
Activities
Courses
Financial aid
SAT/ACT
High school
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Probability of detection
Optimal point on
the ROC curve
depends on reach
capacity and ROI
Probability of false alarm
Predicted
risk Score
6. How is ROC Generated?
Features à Scores à PDF à ROC
GPA
Activities
Courses
Financial aid
SAT/ACT
High school
Model
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Probability of detection
Optimal point on
the ROC curve
depends on reach
capacity and ROI
Probability of false alarm
Predicted
risk Score
7. How is ROC Generated?
Features à Scores à PDF à ROC
Model
Cutoff
threshold
GPA
Activities
Courses
Financial aid
SAT/ACT
High school
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Probability of detection
Optimal point on
the ROC curve
depends on reach
capacity and ROI
Probability of false alarm
Predicted
risk Score
8. Model Performance
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Overlap is a measure of the
model’s ability to separate
between success and failure.
With a strong model you can
be confident of assigning a
particular score to an outcome
category.
With a weaker model, there is
a large amount of overlap, so a
particular score could mean
that an outcome can be either
good or bad with equal
probability.
STRONG
MODEL
WEAK
MODEL
Predicted
risk
score
ROC
9. Parts of a ROC Curve
False
Alarm
Rate
Detec0on
Rate
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Civitas
Model
Random
Ordering
10. Parts of a ROC Curve
False
Alarm
Rate
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Detec0on
Rate
Total Population:
• 10,000 students
• 9,000 continued
• 1,000 did not continue
Point on Line:
• 1,250 students
• 1,125 continued
• 125 did not continue
ROC Information
• Correct identification rate of non-continuing
students = 125/1,250 = 10%
11. Parts of a ROC Curve
False
Alarm
Rate
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Detec0on
Rate
Total Population:
• 10,000 students
• 9,000 continued
• 1,000 did not continue
Point on Line:
• 7,500 students
• 6,750 continued
• 750 did not continue
ROC Information
• Correct identification rate of non-continuing
students = 750/7,500 = 10%
12. Tradeoffs: Without the model, more advisors are needed to reach
more students who will not persist.
False
Alarm
Rate
Detec0on
Rate
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
As you go up and to the right,
you would be reaching out to
more at-risk students (higher
detection rate), but more
interventions require more
advising time and resources
since correct identification
rate of non-continuing
students remains at the same
10%.
13. Model Performance: With the model, the same number of
advisors can reach out to 5X more students who will not persist.
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
Total Population:
• 10,000 students
• 9,000 continued
• 1,000 did not continue
Point on Line:
• 1,250 students
• 1,125 continued
• 125 did not continue
• Correct = 125/1250 = 10.0%
ROC Information:
• 1,250 students
• 650 continued
• 600 did not continue
• Correct identification rate of non-continuing
students = 600/1250 = 48.0%
False
Alarm
Rate
Detec0on
Rate
Civitas
Model
Random
Ordering
~5X
14. Model Evaluation
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
With a stronger
predictive model
• Detection rate improves
• False alarm rate decreases
• Correctness increases at
every student threshold
False
Alarm
Rate
Detec0on
Rate
Civitas
Model
Random
Ordering
15. ACCURACY VS. ROC CURVES
Why is accuracy an incomplete and likely
misleading measure of a predictive model?
16. Accuracy vs. ROC Curves
Case: You use an algorithm to
identify students who are at
risk of not continuing to the
next term.
Following the case study, 10%
of students do not persist.
You test your predictive model
on the data and find that you
made correct predictions 92%
of the time.
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
17. Accuracy vs. ROC Curves
A crackpot scientist tells you,
“I could’ve gotten 90%
accuracy just by predicting
everyone will persist. After all
the math, you gained only
2%?!”
Don’t give up yet!
Your predictive model
is still helpful.
18. Accuracy vs. ROC Curves
You have a team of advisors, and they have time to
reach out to 1,250 students to suggest ways they can
increase their likelihood of persisting.
=
100
students
19. Accuracy vs. ROC Curves
Without the predictive model, you have to pick 1,250
students at random to assist. If 10% of them are
expected to not persist, only 125 students would be
likely to benefit from the intervention.
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
20. Accuracy vs. ROC Curves
With the predictive model, you can choose the
1,250 students by ordering them by the highest
predicted risk score.
The test case reveals 600 of these students are at
risk and would be most likely to benefit from the
right intervention at the right time.
CIVITAS
LEARNING,
INC.
–
CONFIDENTIAL
INFORMATION
21. The ROC Curve Tradeoff
Students most likely to benefit from an intervention
WITHOUT
PREDICTIVE MODEL
WITH
PREDICTIVE MODEL
~5x improvement
22. THANK YOU
VIEW this webinar on-demand on our LinkedIn Page
FOLLOW @CivitasLearning to continue the conversation on Twitter
SHARE comments and ideas for future webinars on the Civitas Learning Space
linkedin.com/company/Civitas-Learning
twitter.com/CivitasLearning
civitaslearningspace.com