08:22:20 From Peter Andresen : How much detail should the answers contain? 08:24:23 From Mirjam Partovi Dilami : Could you give us the plan for week 2 today? 08:24:48 From Sofia Laghouila : When can we expect that the groups will have been formed? 08:25:48 From svend korsgaard : Do what? 08:26:13 From svend korsgaard : ah, i have written. but im not yet in a group 08:26:51 From Kristian Lotzkat : So if we don't have a group, we can just send you a mail and then you'll assign a group to us? 08:43:28 From Marta : red 08:43:29 From Victor von Wachter : red 08:43:30 From Jeppe Karnøe Knudsen : red 08:43:31 From Bia Fonseca : red 08:43:32 From Jens Edelvang-Pejrup : R-E-D 08:43:32 From Kiril : red 08:43:32 From Mikkel : red 08:43:33 From Benjamin Henriksen : red 08:43:34 From Emily : red 08:43:34 From Aske Luja Lehmann Rosted : red 08:43:34 From Julie Kiel Holm : 3rd degree 08:43:35 From Maria : red 08:43:36 From vita : red 08:43:37 From Kristine Marie Løfgren Krighaar : blue 08:43:38 From Daniel Kjær : red 08:43:40 From Victor von Wachter : ALL EXPERTS HERE ;) 08:43:40 From Katriona Gould : red 08:43:42 From Albert Bjerregård Sneppen : Red 08:43:42 From Emma Louise Espersen Knudsen : red 08:43:44 From svend korsgaard : Troels is there anyway to keep seeing your screen while we are in the breakout rooms? 08:43:45 From Jun Jia : red 08:43:45 From Magnus : red 08:43:53 From svend korsgaard : Red 08:45:06 From Peter Andresen : could you give an example og data with uncertainty in y data but no uncertainty in the x data? 08:46:57 From Daniel : What is meant by degrees of freedom? 08:48:37 From Marta : @Peter as far as I understand, you never have no uncertainty, but you can have relatively no uncertainty, fex. if you have a computer recording time with precision of miliseconds or less 08:52:27 From Maria : do we always choose between these 4 models or are there more models for consideration? 08:52:55 From Peter Andresen : Thanks Marta 08:53:01 From Jaime Caballer Revenga : Could you repeat the definition of that Prob(ChiSquare,DoF)? 08:53:05 From Zuzana Moravcova : @Sarah I muted you 08:53:57 From Marcus Winther Dreisler : seems as if most knew it was 3rd degree just from looking at it - I may have missed something but how was that so clear without calculating? 08:54:18 From Jakob : Can this be extended to the case with uncertainity in both X and Y? And is this a common thing? 09:00:59 From Zuzana Moravcova : @Markus: without calculating chi2, you just have to “critically” look at your data points, and check whether your “model” (sin/polynomial of nth degree) is describing your data points well/reasonably enough. e.g. if you have more or less the same # of points above and below your model/function (also the distance from the function is really important, of course) 09:05:22 From Lasse Bonn : comparing this to the least squares problem - what do we do if each model gives similar chi squared, as in the high uncertainty scenario? 09:05:42 From Jakob : Yeah that makes sense, thanks! 09:07:08 From Marcus Nørgaard Weng : In what formula is the uncertainty accounted for in the chi squareds? 09:07:38 From Lasse Bonn : thanks 09:07:38 From Zuzana Moravcova : if I can add to this, if you have so large uncertainties, you cannot really distinguish between different models - so you have to think how you can improve your measurement 09:08:43 From Marcus Nørgaard Weng : Nice, thanks! 09:09:55 From Marcus Winther Dreisler : @Zuzana Yes thank you. I was thinking there might be some intrinsic chi2 consideration to make, but that was just a general consideration on if it looks like it fits….. 09:12:27 From Jakob : How does the Chi square compare to other regression scoring parameters, like R^2, RMSE, RMSEP, RMSECV ? 09:14:06 From Kimi Kreilgaard : How does the Chi2 distribution approach a gaussian when n becomes large? It says so in the book and I think u just said this as well, however to me it looks like it becomes a exponential function chi^n times a gaussian? 09:15:13 From Mikkel Baldtzer Liisberg : For the last exercise (error propagation) we did some fitting, and used the square root of the counts as the uncertainty - can you elaborate on that? 09:22:34 From Peter Andresen : Where does it come from, that the uncertainty is where the delta chi^2 is 1? 09:23:07 From Peter Andresen : Thanks 09:23:45 From Marcus Nørgaard Weng : Can we look at it for a few more seconds if it's so important 09:23:50 From Marcus Nørgaard Weng : So we have time to note it down 09:24:28 From Aske Luja Lehmann Rosted : pretty sure the slides are available on the website 09:24:33 From Marcus Nørgaard Weng : true 09:25:21 From Zuzana Moravcova : both slides and recording are/will be shortly after the lecture online 09:25:44 From Marta : as a pole, id like not to be launched please 09:26:24 From Troels Christian Petersen : Even poles may have astronaut ambitions… 09:27:12 From Marta : fair! 09:28:09 From Norman Pedersen : Its official we are launching Marta now 09:28:32 From Marta : i hope at least i will get astronaut pay then... 09:28:57 From svend korsgaard : i feel like the north pole, but i have no ambition