Timeline for How to fit one distribution to another?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 25, 2015 at 19:50 | comment | added | JimB | @Jagra. You state "I have conjectured that if one takes sufficient observations the data will eventually converge to an exponential distribution." I'm not following. If you give the pdf (which is for a single observation), how can that pdf "converge" to anything with increasing sample size? I can see the distribution converging to something else if a parameter or two goes off to infinity (or approaches some appropriate constants). The pdf you give does not depend on sample size or any specified parameter. What am I missing? | |
| May 25, 2015 at 15:57 | comment | added | Histograms | Doesn't one typically do this by minimizing the KL-Divergence? I guess those intergrals would be too hard to do and the random variates approach is best. | |
| May 25, 2015 at 14:54 | history | edited | J. M.'s missing motivation | edited tags | |
| Sep 18, 2012 at 20:49 | vote | accept | Jagra | ||
| Sep 14, 2012 at 3:02 | history | tweeted | twitter.com/#!/StackMma/status/246443854927130624 | ||
| Sep 14, 2012 at 1:52 | comment | added | Jagra | @whuber -- Thank you for the recommendation. | |
| Sep 13, 2012 at 22:21 | comment | added | whuber | Note that this approach does not solve your stated problem. If you conjecture that the true underlying distribution is exponential, then you ought to consider something like a Maximum Likelihood estimate of the exponential parameter (based on the data you do have) followed by goodness-of-fit testing. | |
| Sep 13, 2012 at 21:23 | comment | added | Jagra | @R.M -- Of course! I should have thought of that. Just late in the day. Thx. | |
| Sep 13, 2012 at 21:19 | answer | added | Sjoerd C. de Vries | timeline score: 2 | |
| Sep 13, 2012 at 21:07 | answer | added | rm -rf♦ | timeline score: 11 | |
| Sep 13, 2012 at 21:01 | comment | added | rm -rf♦ | Since dist is defined easily, it would've been better to do that instead of noting everywhere that the code won't run... it looks much better now :) | |
| Sep 13, 2012 at 20:59 | history | edited | rm -rf♦ | CC BY-SA 3.0 | deleted 101 characters in body; edited tags; edited title |
| Sep 13, 2012 at 20:32 | history | asked | Jagra | CC BY-SA 3.0 |