Stage 2 · Lesson 3 of 8
Conditional probability and base rates
is the chance of something given that something else happened, written P(A|B). Bayes' rule lets you flip it: P(A|B) equals P(B|A) times P(A), divided by P(B). The piece people forget is the base rate, how common A is to begin with.
If 1% of orders are fraud and a flag catches 99% of fraud but also flags 5% of normal orders, most flagged orders are still normal, because there are so many more normal orders. This reasoning appears in fraud, medical tests, spam filters and every classifier you will build.
Common trapconfusing P(A|B) with P(B|A): "most fraud gets flagged" is not the same as "most flagged orders are fraud".
A question to check yourself, an AI practice session and 1 cheat sheet entry come with the full roadmap. Unlock
Resources for this stage
Start with· follow one of these
- VideoFreeStatQuest with Josh Starmer
StatQuest · Beginner
Free: short, clear videos on each concept, from the normal distribution to p-values and power
- CourseFreeKhan Academy Statistics and Probability
Khan Academy · Beginner
Free: the probability, sampling distributions, confidence intervals and significance tests units, with practice
Practise
Keep handy
Also worth your time
Interactive visuals for probability, distributions, inference and regression: play with each one