Thursday, November 27, 2014

Higgs hunters: a weird citizen science project

The ATLAS Kaggle Higgs contest made a lot of sense. Today, some NYU+Oxford folks from ATLAS – working with the Zooniverse – have presented something similar. Except that I have no idea how I could usefully help them:
It's particle-hunting season! NYU scientists launch Higgs Hunters Project (Phys.ORG)
The press release above promotes the website
where you may sign up (with Zooniverse) and start to classify collisions.

This new ATLAS picture isn't really from the contest but it's fun: it's showing a cosmic ray particle going through the magnetic fields in the detector.

What you will be asked isn't to write a clever computer program, like the contestants in the Kaggle competitions. Instead, you will visually evaluate some pictures of ATLAS collisions.

The ATLAS folks have probably run out of computers so they need to employ humans to do the job! Can you perform a few billion mathematical operations in a second? Or do you think that you may beat the computers just by your natural skills, intuition, and good eyesight?

Your task is to look for "off-center vertices" in the pictures – places away from the beam (from the center of the picture) where several or many trajectories of particles intersect. Some of those vertices could be decays of new "somewhat long-lived" particles.

The evaluation of a picture isn't a task in which the computers are maximally talented. However, I simply can't imagine how I could possibly do better than even rather simple programs that look at these pictures. Why isn't this contest about writing a program to recognize the patterns, like the Kaggle contest, instead of this manual work?

Moreover, I don't understand how the candidate collisions with possible "off-center vertices" have been selected. I hope that ATLAS isn't offering you random collisions taken from the list of trillions of collisions in the detector because that would be rather hopeless, indeed.

I've registered but I think that this is the end of my playful times on this server. I have no idea what to do with it. If you get further or decide that it's interesting or that you can even discover something that the experimenters can't, please, let me know!


  1. Dear Lubos, by double slit experiment are you suggesting arguments presented in Feynman's Lectures on Physics ? I have never learned Bel's theorem, Feynman's arguments were very convincing for me. I think a physicist should be satisfied with this and shouldn't look for a complete mathematical proof because you can never have a rigorous proof for all kind theories therefore it is pointless. Also I belive even if we didn't know relativity quantum mechanics should be accepted because it will be enormously more simple than a highly non-local theory you need to figure it out.

  2. Dear John, I mean my arguments similar to his ideas about the double slit experiment - all wisdom of QM is encoded in that experiment, as Feynman would say - but reformulated as a provable inequality analogous to Bell's inequality.

  3. Crowdsourcing should offer rewards (a penny/image identified) or recognition (ranking by number of images identified). That encourages obsessive-compulsive, testosterone-fired competition re SETI online.

    Why steal high intelligence to do assembly line work? Internationally enlist prisoners. Trade accurate productivity against time served, perhaps 1000 images/(day deleted from remaining sentence). If they are inaccurate, one day added/20 incorrect submissions

  4. yes, a Kaggle clone contest would be better---just get a good machine learning algorithm/pattern recognition, and turn it loose with one of Gregory Chudnovsky's DIY supercomputers ;)

  5. Yup, thanks.A fun video.

    BTW all married climate skeptics should check that their wives are not gullible. DeSmogBlog started its campaign to break the marriage of all the "climate deniers" on their black list. Dana Bate

    is the first told about her nightmare with her denier husband who dares to question the climate apocalypse which is just like if he questions the health impact of smoking.

    Behind her gleaming smile and professional friendliness hides a very American nightmare.


    Does this brilliant and charismatic reporter actually know how her husband has been making his money?

    Is there a hidden meaning in the titles and themes of her books? Does her beloved Roger deserve a second bite of the apple? Are we being asked for our forgiveness?

    Or should we forever remind Roger about the secret supper clubs he ran, advising tobacco and oil companies on how to convince the public that the science linking smoking to cancer and fossil fuel burning to climate change should be considered questionable?

    You can't make this up. I don't remember that communists would ever go "this deep" into the families of the dissidents.

  6. Nicely done video that summarizes it all pretty well.

    OTOH, I am convinced that, these days, if you mention double slit experiment to well-informed non-experts, the reaction would be: "Meh, photons/electrons sometimes behave as particles and sometimes as waves, big deal."

    But if you mention:

    you will likely get a reaction: "Oh, that is so strange"

    So there is an aspect of fashion to what phenomena get media attention, which in itself is nothing new or surprising.

  7. You are free to mumble nonsense but not to state it as "my opinion". FYI I hope that the Front National will not come to power in France because at its core it is a party composed of xenophobes, homophobes, anti-semites, revisionists, religious nuts (hello Shannon) whose founders included collabos and nostalgic of Vichy. A party that rejects the French Revolution and its legacy and whose main economic doctrine is that of protectionism. I could go on but that's the essence of it.

  8. The problem is that the meanings of terms like 'conservative' and 'liberal' have completely changed over the years to the point they're now ambiguous.

    'Liberal' used to mean libertarian; now it means the opposite.

    I've lost track of what 'conservative' is supposed to mean. Some people think it means keeping government small (usually with the exception of the military, which, despite being the very quintessence of government, for some reason gets a pass) but not allowing people to ingest the wrong plants.

    The fact is -- most people who bandy these terms about have not studied history or economics or politics to any extent or devoted much time to thinking about the issues; but if enough of them start using the word 'liberal' to mean being in favor of a large, redistributive (from the middle to the top and bottom) corporatist state, then that's what it comes to mean. Actually, 'neoliberal' is a more accurate word to describe this type of ideology. Unfortunately, most 'conservatives' in leadership positions also embrace 'neoliberal' ideology.

  9. Is Quantum Entanglement Real?

  10. Just see the reaction on peoples face when you tell them that a particle can exist in two or more places at the same time.

  11. I see there are other 20 tasks seeking volunteers. I tested randomly in the 'Floating Forest' and is also visually identify certain images on maps.
    It seems evident that search patterns are suitable for MachineLearning.

    Maybe this is the "supervised" part in a supervised Machine Learning: Marking classes in few thousand of examples, and then the computer learn from it and classified million. This would make sense. Maybe it's in the fine print:)

  12. User Agreement "..The major goal for this project is for the analyzed data to be available to the researchers for use, modification and redistribution in order to further scientific research."

    My hypothesis is in that "further" :)

  13. I wrote 'well-informed non-experts' which would mean you, not your grandma.

  14. I hope something like UKIP becomes the new normal here in the United States too.

  15. Dear Lubos, you misinterpret what FN is and what it aims at. I won't talk about UKIP or FPÖ because I don't know them enough.
    But FN is for me very clear..
    The FN of father Le Pen who created it was a classical far right party. This inculdes but is not limited to :
    vague antisionisme, preference for a strong state; support for/from religion (especially traditionnalists), harsh justice especially against strangers, strong anti islamism, very strong anti establishment stance, strong anti communism. Main slogan : "La France aux Français" e.g France belongs to French and French were those who lived in France for (many) generations. For the rest it was "Ausländer raus".
    Economy was never mentionned like something important. If anything father Le Pen was a liberal capitalist.
    With this very clear cut definition, the electoral pool was anywhere between 5% and 10 %. Never less but no potential for significantly more.
    One agreed or not with this program or parts of it but it was irrelevant because this party was never going to get enough voters to count.
    The FN of the daughter Le Pen is something entirely different altogether.
    Her strategy was very simple - tap other electoral pools in order to break the limit of 10% and become a contender for power.
    Makes sense.
    She didn't change much the nationalist topics of her father. Well she toned strongly down the anti communism which characterised her father but basically kept everything else around the "La France aux Français" theme.
    The internal revolution happened on economy policies. The liberal capitalism became the devil and what was adopted was a marxist point of view.
    Capital is evil. Free trade is evil. Banks are evil. The poor French are oppressed and exploited by evil multinational companies and by the evil imperialists which are even more dangerous and detrimental than arab immigrants (but to be sure, the latter are of course still evil N°2).
    France has to construct a duty maginot line that would tax all this evil exporting its goods to France so that it stops. France has to close its frontiers and get out of the Euro. A minimum salary (1500€) has to be guaranteed. Etc.
    She transformed the FN in an authentic national socialist party.
    Of course the electoral results followed. She kept broadly the 10% of the old nationalist core but added the whole far left spectrum which was delighted by her "economic" speeches.
    Amusingly the leaders of the far left (communists, altermondialists, ecologists, left front) are still attacking the FN as an evil far right party but their voters massively deserted to FN.
    FN is today, and by far, the strongest party in the worker and state employee segment where it replaced the communists.
    Some of the most radical far left politicians also deserted and switched to FN.
    So while I never considered the father Le Pen's FN as dangerous (like virtually 100% of the PC media), I consider the daughter Le Pen's national socialist FN as very dangerous.
    If they ever came to power and applied their marxist economical program, it would be a disaster compared to which the Greek episode would look like a harmless amusement.
    You see1948 Lubos ? Well Le Pen would be 48 squared.

  16. Dear Tom, I have never claimed that FN was a pro-market party or something like that.

    On the other hand, I don't quite believe that it would be (significantly?) more anti-market etc. than the Socialists now - and maybe even than the Sarkozy types.

  17. I can't see how pointing out PC delusions can be regarded as ''promoting'' these very delusions. It's like claiming your antifeminist comments make the feminist ''opinions more visible''.

  18. Philip wants to make his understanding a universal truth. Typical.

  19. Bryan was not a socialist, but, yeah, populist-leaning. At least the ''progressives'' faction of the Republicans (Roosevelt in 1912) were definitely to the left of Democrats. And yes, Th. Roosevelt was no socialist either.

    Also, let us not ignore the fact that even FDR actually used anti-socialist/minimal state rhetoric in his 1932 campaign (,_1932#Campaign), though he went on to pursue interventionist policies in office. So I'd say 1930s were the most important period concerning the re-alignment ''Reps - more laissez faire/Dems - pro-state interventionist''.

  20. It's typical of everyone, Shannon. Lubos and others do it with respect to ,e.g., quantum mechanics. Surely the only question is the accuracy of one's understanding and whether or not it is even possible that it is approaching a universal truth.

  21. Lucy In the Mirror PairNov 28, 2014, 3:53:00 PM

    Motl, I am in love with you for the true hero that slowly but unceasingly you simply are; of Science near the midnight of its darkest hour.

  22. It is called scanning. I have scanned thousand of pictures back in the 70's. Computers are fine if you know what you are looking for. I guess they are fishing . They are presenting frames that the computer has thought belong to one interaction time, one event. We must not forget that there are 40 or fifty interactions in the region and the computer programs have decided on this picture.

    I did the free "tutorial", and did not register. They are presenting frames where the computer program has decided there are more than one vertices belonging to the interaction. What put me off are free floating tracks in some of them. If I do not know how the decision was made to associate these free floaters with the main vertex I do not want to waste my time. It seems to me possible they are mislabeled tracks from another interaction in the region, just from charge conservation. We are asked to comb the bath water from the computer decisions in case there is a baby in it.

  23. Exactly, Anna. To summarize your comment in a more vague way, there are lots of difficult technical features concerning how the pictures were selected and what you should do with them that don't have clear answers and that must quickly convince a user that he would be wasting her time if she were playing with that for hours. It's hard to see "how well you are doing", "how much chance of seeing something interesting" you really have, and why it wasn't seen before.

  24. Yes I understood what you said. I was just talking about general people. I just thought it is interesting the different reactions you get from regular people, something like asking for your experience.

  25. Lubos - Of course 'words mean things' but, more importantly, 'people mean things'! Simply appending the word 'justice' to some other word or phrase does not mean that the concept of justice has been extended in some way. What is meant (i.e. - what people mean) by 'social justice' is more akin to some primitive and even childish notion of 'fairness'. All of these '(-----fill in the blank-----) justice ' constructions are intended to deceive, and they have had remarkable success, not least that you and Werdna have found yourselves in such a dispute.
    Wedna is an intelligent and knowledgable commenter who writes in good faith and I'm sure that no insult or condescension was intended.
    I do have one question - When such a 'sick generalization' results in actual injustice could one then say it was the opposite of justice?

  26. They would be incredulous, and rightly so!

  27. Dear Philip, it isn't the same thing because I explain what is counterproductive about feminist views and do so in my language.

    On the other hand, you have been repeatedly trying to force me to adopt the PC vocabulary (that you seem to share), and perhaps not the vocabulary.

    Sorry, I won't do that.

  28. Dear RAF, I agree with Werdna - and you - on the substance of the political argument, and have respect for that.

    Another way to explain the reason why I think it's wrong to call them "opposite" is that historically, some of the people who were fighting for justice - what we would agree is justice - were wanting "social justice" as well.

    You present them as being in tension with each other but it's at least equally natural to present them as "similar things".

    I do have one question - When such a 'sick generalization' results in actual injustice could one then say it was the opposite of justice?

    No, of course that it is not enough. It's my point. If one example of "social justice" results in "injustice", it just implies that "social justice" and "injustice" are not exactly equivalent - they're not synonyma.

    To show that they're "opposite" to each other, it would have to be that in *every* case, what "social justice" implies should be "injustice" - "social justice" and "justice" would be opposite to one another is "social justice" were (exactly) the same thing as "injustice". ;-)

  29. viewed in "slice" frame a lot of them look like "beam gas" interactions that the computer has caught with some random mixed from other interactions tracks. the one you have in the picture is interesting, as it does not look beam gas.

    They are offering high p_t missing events with with vertices "thought" by the computer. The one in the above picture could be a cosmic neutrino hitting a detector atom, at the center :) .

  30. Dear Luboš,

    "I am planning a blog post about the double slit experiment that not only shows this general "Bell's theorem" but makes it clear that all of its parts were fully appreciated decades before Bell."

    Jolly good!

    It seems to me that this two-slit 'tweet' of yours could nicely set the cat among the pigeons and cook a whole lot of jive turkeys' geese for them.

    I can't wait to smell the roast and watch the shit hit the fan. :)

    Is there any chance you could make this poultry classic of yours available as a pdf?

    "♬ Four and twenty blackbirds baked in a pie
    Wasn't that a dainty dish to set before the king...♩♫

    Happy days! :)

    P.S. It would increase the pleasurable effect for me if the deadly poisonous sting you deliver were to be viciously painful and all concentrated in the tail of your piece, if you get my drift. No pressure though. Tee hee! :)

  31. Popular blog Gizmodo, part of the far left wing tabloid Gawker media conglomerate, is today pushing a "physics is phucked" article relating relativity and dark matter/energy to the simulation hypothesis:

    Odd unknowns = glitches in the Matrix.

  32. This post is so informative and makes a very nice image on
    the topic in my mind. It is the first time I visit your blog, but I was
    extremely impressed to see blog about quantum entanglement. Keep posting as I am gonna come to read it everyday!