Statistical Analysis of Massive Data Streams Proceedings of a Workshop (2004) / Chapter Skim
Currently Skimming:
Pages 92-113
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 92... ...
What are the ~ : ~ ~ ~ that control their behavior at the most basic level So, I am doing particle physics which you just heard about, and my brief summary of what we tried to do is answer the following questions. What are the elementary constituents of matter, and what are the basic forces that control their behavior at the most basic level?
|
From page 93... ...
So, why is this point particle, that is infinitely small, have a mass? So, there are some pretty big, weighty questions that haven't been answered by the standard model.
|
From page 94... ...
Each step in the data reduction is a statistical selection process Think of a particle physics detector as a "Smart Devices You can think of a big particle physics detector as a smart device. It is a realtime smart device doing real-time analysis.
|
From page 95... ...
"Almost 100 physicists an~o,., professional statisticians ~~ gathered at Durham's IPPP in March to discuss statistical techniques in particle physics." Okay, there is an effort to rectify this. Here is a headline from what ~ will call a trade magazine, the CERN Courier, a standard popular magazine about a business: "Physicists and Statisticians Get Technical in Durham." There was a big conference in Durham last March on applying statistical techniques to particle and astrophysics data.
|
From page 96... ...
First, a little bit of what we do. We talked about how we collide protons, for example, and look at what comes pouring out into our detector event by event, and then we statistically analyze that looking for new physics or known physics.
|
From page 97... ...
What we do, this is what we see in the detector, and we have to statistically analyze that to figure out what went on. Then we will make an invariant mass distribution of these two photon signals over a large statistical event to look for a bump.
|
From page 98... ...
~ " determine what happened - A I: ~ if- ~ 1 complex pattern recognition problem - ~= (with lots of cluster finding) :~ .;_ Statistically analyze the events and look for "physics" i 8 ~~ 1 U.: So, we have sort of a two-step problem.
|
From page 99... ...
~4 IT.: It is not just those bumps we look for. Sometimes, what we are looking for are differences from the standard model that are a little bit more subtle.
|
From page 100... ...
The "Black Box" nature of Neural Networks worries many in the field A group at Rice developed PDE (Probability Density Estimation) as an understandable alternative.
|
From page 101... ...
describes a surface in n dimensions. You form the signal and background surfaces using the training data You can think of the surfaces as just smoothed representations of the training data This function describes a surface and end dimensions, and you form the signal ant!
|
From page 102... ...
102 It reflects the covariance of the data. You construct a transformation, so that one class of the data has a covariance matrix that is a unit matrix, ant!
|
From page 103... ...
103 distribution, and here is a background.
|
From page 104... ...
Where x is a reference point (we are currently using the mean) We have modified that original method a little bit.
|
From page 105... ...
So, we have this standard model that works beautifully and explains all the known data. ~ think if you made a poll of the particle physics group, nobody would believe that the standard model is right.
|
From page 106... ...
So, how do you possibly know -and changing those parameters changes what you will see in the detector. ~ t~` -win irk '~ Nutshell ~ We have a well defined standard model.
|
From page 107... ...
So, the typical physics analysis done in a particle physics experiment is done at I.5 on this scale, and we would like to be up here at 6.0, searching through the data, looking for new things. NI(~.~L NT(~LiV8ti(-~l Another related issue: How do we q..antify the "interestingness" of a few strange extents a poster~ori?
|
From page 108... ...
108 ~ Sleuth Step ~ Exclusive final states ·3 ~ ' ~ a: :" ' ¢: -.~ :3 a: " 2 a..
|
From page 109... ...
109 Algorithm Variable transformation ,,,,,.,,,,,,, ,,,, ,,,,,, ,,,,,, ,,,,,,,,,, ,,,,,, ,,,,,,,, ,,,,, ,,,,,,,,,,,,,,, ,, ,,,,,,,.,,, ~ , ............................... We begins by applyir~g a varial~le transfor'~at~on that Intakes tl,e hacXgr:~` Ed distrit)
|
From page 110... ...
110 Algorithm Variable transformation ,,,, ,,,,,,,,,,.,,,,,,,,., , ,, ,,—, ,,,,,,, ,,,, ,,,,—, The tra~sfo~~`at~o~ maps the signal region into the upper r~gl~thand corner of the unit box .
|
From page 111... ...
This method was testes} on known physics. For example, the search for the top quark, which was a big discovery, was reproduced using this.
|
From page 112... ...
,,,,,,,, ,,,,,,,,,,,__, ,,,,,,,,,,— _,,,,, ,,,,,.,,,, ,,,, ,, ,, —, ~ ,, ,,,,,,,,,,,,,,,,,,,,.,,,—,......................... Particle Physics presents a number challenges of interest - ~ have only scratched the surface Q We have large data streams that we must search in real time and offline for signals of (possibly unanticipated)
|
From page 113... ...
113 some distribution that is fine with exponential pay off. You have two events way out here, far away from the standard model.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.