ML Wiki
Machine Learning Wiki - A collection of ML concepts, algorithms, and resources.

Point Estimate

Points Estimates

An estimate is a point estimate when we use only one certain value to describe what we think about the whole population [https://en.wikipedia.org/wiki/Point_estimation]

Examples

  • $\bar{x}$ - sample mean, $\mu$ - population mean, $\bar{x}$ estimates $\mu$
  • $\hat{p}$ - estimate of proportion, $p$ - true proportion
  • $s$ - sample Standard Deviation, $\sigma$ - std of population
  • etc

Point estimates are not necessarily good, but they tend to become better as we collect more data

Standard Error

Standard Error - a measure of uncertainty associated with this point estimate

Confidence Intervals

A point estimate - is a single plausible value for a parameter

  • but one point is rarely enough - we’d like to have a range of plausible values
  • so we typically use Confidence Intervals

See Also

Sources

  • OpenIntro Statistics (book)
  • http://onlinestatbook.com/2/glossary/point_estimate.html
  • http://en.wikibooks.org/wiki/Statistics/Point_Estimates