Scribd
Upload
73 views

Sets: Medical Testing Example: 1 Example Using Set Theory

This document discusses using set theory and medical testing to determine if a person has a syndrome. It defines sets for the total population in a clinical trial (X), those who have the syndrome (S), and those who do not (H). A test is used to determine who tests positive (P) or negative (N) for the syndrome. The goal is for S to equal P (true positives) and H to equal N (true negatives), but false positives, negatives, and rates are also discussed. Cardinality is analyzed to determine proportions with or without the syndrome and ideal true/false positive and negative rates.

Uploaded by

Yandi Layadi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
73 views

Sets: Medical Testing Example: 1 Example Using Set Theory

This document discusses using set theory and medical testing to determine if a person has a syndrome. It defines sets for the total population in a clinical trial (X), those who have the syndrome (S), and those who do not (H). A test is used to determine who tests positive (P) or negative (N) for the syndrome. The goal is for S to equal P (true positives) and H to equal N (true negatives), but false positives, negatives, and rates are also discussed. Cardinality is analyzed to determine proportions with or without the syndrome and ideal true/false positive and negative rates.

Uploaded by

Yandi Layadi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Data Science Math Skills

Paul Bendich and Daniel Egger


Duke University

Sets: Medical Testing Example


Video companion

1 Example using set theory


VBS: “very bad syndrome”

X = set of people in a clinical trial

S = {x ∈ X : x has VBS}
H = {x ∈ X : x does not have VBS}

X =S∪H (you either have VBS or you don’t)


S∩H =∅ (no one both has and doesn’t have it)

Point of medial testing to figure out whether a person is in S or in H

2 Test
P = {x ∈ X : x tests positive for VBS}
N = {x ∈ X : x tests negative for VBS}

P ∪N =X (you either test positive or negative)


P ∩N =∅ (no one tests both positive and negative)

In a perfect world, S would equal P —the sick people would always test positive, and H
would equal N —the healthy people would always test negative.
...but this is not always the case.

S∩P H ∩N S∩N H ∩P
true positive true negative false negative false positive

1
Data Science Math Skills
Paul Bendich and Daniel Egger
Duke University

3 Cardinality
|S|
|X|
= proportion of people in the study who do genuinely have VBS
|H|
|X|
= proportion of people in the study without VBS
|S| |H|
|X|
+ |X|
=1

|S∩P |
|S|
true positive rate would like to be close to 1
|H∩P |
|H|
false positive rate would like to be as small as possible
|S∩N |
|S|
false negative rate would like to be as small as possible
|H∩N |
|H|
true negative rate would like to be close to 1

You might also like