tag:blogger.com,1999:blog-4453805095942812863.post8426891890885835065..comments2024-02-23T02:14:03.787-08:00Comments on Surprise and Coincidence - musings from the long tail: Exponential weighted averages with irregular samplingTed Dunning ... apparently Bayesianhttp://www.blogger.com/profile/02498665124454933570noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-4453805095942812863.post-91037725929609755832016-07-31T15:15:10.788-07:002016-07-31T15:15:10.788-07:00very useful article
thanks
this link
http://www...very useful article <br />thanks <br />this link <br />http://www.thalesians.com/archive/public/academic/finance/papers/Zumbach_2000.pdf<br />is really great, too<br />thanksAnonymoushttps://www.blogger.com/profile/00815025625155764930noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-32542446725213920362014-09-04T18:01:55.150-07:002014-09-04T18:01:55.150-07:00Excellent paper!Excellent paper!Ted Dunning ... apparently Bayesianhttps://www.blogger.com/profile/02498665124454933570noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-59920900453620024062014-09-04T16:48:04.851-07:002014-09-04T16:48:04.851-07:00Just as an FYI to anyone ending up here, there'...Just as an FYI to anyone ending up here, there's a truly excellent paper with many useful online time-series operators for inhomogeneous data:<br /><br />http://www.thalesians.com/archive/public/academic/finance/papers/Zumbach_2000.pdfone_fell_swoophttps://www.blogger.com/profile/04415486703354218677noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-40504478547848665432012-03-16T21:34:06.845-07:002012-03-16T21:34:06.845-07:00That worked, thank you much. Your CAPTCHA is nearl...That worked, thank you much. Your CAPTCHA is nearly impossible, BTW. Or maybe I'm a robot and it is just doing its job. :)Nathanhttps://www.blogger.com/profile/08660694035380786136noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-72022106858275460772012-03-16T17:26:46.618-07:002012-03-16T17:26:46.618-07:00And, by the way, thanks for noticing that. There ...And, by the way, thanks for noticing that. There was also a missing minus sign on an earlier expression that I found as a result of looking for the problem you spotted.<br /><br />Let me know if you have any other problems with this.Ted Dunning ... apparently Bayesianhttps://www.blogger.com/profile/02498665124454933570noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-89753675870501355772012-03-16T17:25:14.602-07:002012-03-16T17:25:14.602-07:00Nathan,
The issue is that I dropped a division by...Nathan,<br /><br />The issue is that I dropped a division by α in the expression for $\pi$. For the example, α=5 so for the second line, $\pi = \exp{-(21.54637-11.35718) \over 5} = e^{-2.04} = 0.1303101$Ted Dunning ... apparently Bayesianhttps://www.blogger.com/profile/02498665124454933570noreply@blogger.comtag:blogger.com,1999:blog-4453805095942812863.post-63345238334838514912012-03-15T23:59:28.810-07:002012-03-15T23:59:28.810-07:00Ted, pardon my bad math, but I dont understand how...Ted, pardon my bad math, but I dont understand how you are calculating pi in your example. I can't get to .13031 no matter what I do. ln .13031 is -2 ish, which doesn't seem to come from t1 - t0, which is about 10 - what am I missing?<br /><br />Thanks,<br /><br />NathanNathanhttps://www.blogger.com/profile/08660694035380786136noreply@blogger.com