| (2 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
Biased or unbiased? | Biased or unbiased? | ||
| − | E[1/x] = integral(<math>\lambda e^{-\lambda x}</math> * (1/x) dx is undefined | + | *E[1/x] = integral(<math>\lambda e^{-\lambda x}</math> * (1/x)) dx * this integral is undefined |
| − | + | *therefore biased... because E[estimator] does not equal estimator | |
| − | therefore biased because E[estimator] does not equal estimator | + | |
Latest revision as of 14:48, 10 November 2008
Biased or unbiased?
- E[1/x] = integral($ \lambda e^{-\lambda x} $ * (1/x)) dx * this integral is undefined
- therefore biased... because E[estimator] does not equal estimator
