Timeline for How is this minimum variance worked out for this importance sampling estimator?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 25, 2018 at 1:35 | vote | accept | Lerner Zhang | ||
| Jan 25, 2018 at 1:29 | comment | added | shimao | It is because the derivative of $1/f(\epsilon)$ is $-\frac{f'(\epsilon)}{f(\epsilon)^2}$ by the chain rule, where in this case, $f(\epsilon) = q(x) + \epsilon \eta(x)$ | |
| Jan 25, 2018 at 1:24 | comment | added | Lerner Zhang | Thanks for your derivation. But how does the first limitation becomes the second? It seems that the divisor is squared unknowingly... | |
| Jan 24, 2018 at 14:55 | history | edited | shimao | CC BY-SA 3.0 |
edited body
|
| Jan 24, 2018 at 14:49 | history | edited | shimao | CC BY-SA 3.0 |
edited body
|
| Jan 24, 2018 at 14:44 | history | answered | shimao | CC BY-SA 3.0 |