Logs, Tails, Long Tails – Ryan Moulton's Articles
moultano.wordpress.com · 1,431 words · saved by 3 readers
Learn the shape of the logs of probability distributions, and you’ll be able to guess how they behave.
Gaussian Guessing Game Guess which of these graphs shows the normal/gaussian distribution. A B C D The correct answer is A. The rest are the logistic, cauchy, and beta distributions respectively. Even if you picked it out correctly, it isn’t easy to figure out. All of them look like “bell curves,” they are all symmetric, they all taper off at what looks like a similar rate. In that case, why do we care about the difference between them? Why don’t we just pick whichever one is most mathematically convenient and not worry about it? The difference lies in their tails, the parts of the graph where
saved by
related reading
- Searching for outliers | benkuhn.netbenkuhn.net
- Six (and a half) intuitions for KL divergence — LessWronglesswrong.com
- How much do you believe your results? — LessWronglesswrong.com
- What is Fat Tail? | TDS Archivemedium.com
- How to Transform Data to Better Fit The Normal Distribution - MachineLearningMastery.commachinelearningmastery.com
- 8 Bayes’ Theorem | Odds & Endsjonathanweisberg.org
- Understanding Momentsgregorygundersen.com
- A Maximum Entropy Intuition for Fundamental Statistical Distributions | long intuitionlongintuition.com
- Everything Is Logarithmsalexkritchevsky.com
- Sub-Gaussian distribution - Wikipediaen.wikipedia.org
- The Value of Probabilistic Thinking: Spies, Crime, and Lightning Strikesfs.blog
- Why the tails come apart — LessWronglesswrong.com