Statistical Inference I

on YouTube ↗56m

" summary

Defines a statistical model, loss and risk, then compares three binomial estimators delta0 = X/n, delta1 = (X+3)/n, delta2 = (X+3)/(n+6) and plots MSE vs θ to show delta1 is inadmissible while delta0 and delta2 trade bias for variance. Explains bias–variance decomposition, Bayes vs minimax criteria (uniform prior yields Bayes estimator = (X+1)/(n+2) for squared loss), notes Bayes-optimal rule is the posterior mean, discusses empirical/hierarchical Bayes and practical computational/prior elicitation issues.

AI-written summary · about

" chapters (15)

1. 0:00Introduction to estimation
2. 0:54Statistical models
3. 3:33Estimators and statistics
4. 5:32Loss and risk functions
5. 7:26Understanding risk functions
6. 9:16Evaluating estimators
7. 15:30Bayes and minimax risk
8. 18:39Minimax risk properties
9. 25:24Unbiased estimators
10. 27:37Bias-variance tradeoff
11. 32:31Bayesian inference
12. 41:10Choosing priors
13. 44:48Randomized estimators
14. 47:34Empirical and hierarchical Bayes
15. 53:03Using non-informative priors