Shelf
Statistics · Beginner · 5 h
Probability and distributions
Probability rules, conditional probability and Bayes, and the binomial, Poisson and normal distributions you meet in analytics and interviews.
Why employers care: Interviewers test probability directly, and every confidence interval, A/B test and anomaly alert rests on these distributions.
Lessons, in order
- Probability basics: rules and conditional probabilityCompute probabilities, conditional probabilities and expected values from 41,188 real bank marketing calls, test independence by hand, and check the classic interview puzzles by simulation.11 min
- Bayes theorem explained with real examplesWork out why a fraud detector that is 99% accurate is wrong most of the time it fires, verify it by simulating a million transactions, and update real bank marketing probabilities one piece of evidence at a time.11 min
- Binomial and Poisson distributions with examplesModel sign-ups from a batch of calls as binomial and taxi pickups per minute as Poisson, compute probabilities with scipy.stats, and catch overdispersion on real data before it fools your forecast.11 min
- Normal distribution and z-scoresFit normal curves to penguin body mass, check the fit with Q-Q plots, compute areas and percentiles with scipy.stats.norm, and see why taxi fares need a log-normal and why z-score outlier rules fail on skewed data.12 min