Thinking like a Bayesian
summary
Claim: Bayes' rule isn’t mainly a formula to plug numbers into but a source of intuitive habits for updating beliefs. Describes how modern spam filters use Bayesian updating to raise spam probability when certain words recur, and gives a concrete social example where the right move is to ask what you'd expect if the hypothesis were wrong rather than jump to confirmation. Emphasizes comparing predicted observations under competing hypotheses before boosting confidence.
AI-written summary · about
related talks
25mWhy Be Bayesian If I'm a Frequentist?
25mScaling Bayesian Experimentation
29mHow to use Data Science Superpowers in real life, a Bayesian perspective
31mBias Meets Bayes: A Bayesian Perspective on Improving Model Fairness
43mToward Human-Centered Machine Learning
01h 30mBayBE: A Bayesian Back End for Experimental Planning in the Low-To-No-Data Regime
31mFrequency and Bayesian Approaches to Estimating TPR with Partial Labels
24mThe Scientific Method of Experimentation