/13 min read
Parameter Estimation
What it means to estimate a parameter, and a map of the six workhorse estimators — LS, MVUE, MLE, LMMSE, MMSE, MAP — chosen by asking what you actually know.
- estimation-theory
- foundations
Topic
Estimation and detection from noisy observations — how to infer parameters, bound how well anyone could, and adapt when the statistics move.
What it means to estimate a parameter, and a map of the six workhorse estimators — LS, MVUE, MLE, LMMSE, MMSE, MAP — chosen by asking what you actually know.
The parameter is deterministic and unknown — minimum variance unbiased estimators, the Cramér–Rao bound, and maximum likelihood.
The parameter is a random variable with a prior — MMSE and MAP estimators, and what the posterior buys you over a point estimate.