Burn-in and thinning — why?
mediumAnswer
- Burn-in: discard initial iterations before the chain reaches stationarity.
- Typical: 10-50% of chain.
- Thinning: keep every k-th sample to reduce autocorrelation.
- Modern view: thinning wastes information unless memory is tight — better to use all samples with autocorrelation-aware summaries.
- Use ESS-based decisions rather than fixed burn-in / thinning rules.
- NUTS + adaptive warmup replaces manual burn-in in practice.
Check yourself — multiple choice
- Random
- Burn-in: drop pre-stationary samples; thinning: reduce autocorrelation storage. Modern: use all samples with ESS-aware summaries; NUTS auto-warmup replaces manual burn-in
- Never needed
- Same as ESS
Burn-in: discard pre-stationary; thinning: memory tradeoff; NUTS warmup preferred.
#bayesian#mcmc
Practise Statistics Fundamentals
215 interview questions in this topic.