Forgot your password?
typodupeerror
Software Space Technology

AI Astronomer Aids Effort To Analyze Galaxies 40

Posted by timothy
from the kindly-disprove-loyalty-to-alien-invaders dept.
kkleiner writes "Scientists are teaching an artificial intelligence how to classify galaxies imaged by telescopes like the Hubble. Manda Banerji at the University of Cambridge, along with researchers at University College London, Johns Hopkins, and elsewhere, has succeeded in getting the program to agree with human analysis at an impressive rate of more than 90%. Banerji used data from Galaxy Zoo, a massive online project that has used more than 250,000 volunteers to analyze more than 60 million galaxies. The new automated astronomer will help with even larger analytical projects on the horizon, taking care of trivial classifications and leaving the tough cases to humans."
This discussion has been archived. No new comments can be posted.

AI Astronomer Aids Effort To Analyze Galaxies

Comments Filter:
  • better than humans? (Score:4, Interesting)

    by samwise098 (1463055) on Wednesday June 09, 2010 @04:07PM (#32515426)
    I wonder how this program compares to a human doing the same job If given the same "training" I wonder how many humans would get a 90% agreement rate looking at the same data.
    • Re: (Score:1, Redundant)

      Well exactly. This isn't so much an "AI" as it is a script that has output depending on its input.

      Essentially, any "errors" the computer makes would be an issue with setting up a debugger and seeing why it chose that selection. Then it falls into two categories: It spotted things humans missed when classifying or it has faulty programming that is creating false positives or missing data.

      I'd hardly consider that "AI".

    • Re: (Score:3, Interesting)

      by Locke2005 (849178)
      Sense the "training" consists of being shown hundreds of thousands of galaxies along with their classifications, I doubt that many humans would live long enough to be given the same "training"...
  • Name (Score:5, Funny)

    by kellyb9 (954229) on Wednesday June 09, 2010 @04:10PM (#32515458)
    I would think with a name like Al Astronmer your career choices would be limiting. I guess I was right.
  • Original paper (Score:5, Informative)

    by JoshuaZ (1134087) on Wednesday June 09, 2010 @04:17PM (#32515536) Homepage
    The paper discussing this work is http://arxiv.org/abs/0908.2033 [arxiv.org]. They appear to be using a pretty standard neural network approach (disclaimer: I don't have much background in neural nets at all. I'm just going off of how they were described in the last class I took that discussed them.) This is part of a very general pattern where programs have done a lot of work that we would think could only be done by people. Other examples include the computerized proof of the Robbins conjecturehttp://en.wikipedia.org/wiki/Robbins_conjecture [wikipedia.org]. TFA lists a few examples as well which are in more applied areas.
  • I wonder if it has any Tauntauns!
  • by Locke2005 (849178) on Wednesday June 09, 2010 @04:28PM (#32515660)
    If this technologies works for classifying galaxies, perhaps next we could put it to work classifying porn on the web!
    • Re: (Score:1, Funny)

      by Anonymous Coward

      For the love of God, delete this post! This comment is the reason Skynet starts down the logic path of removing humans!

      - John Connor

    • by Adambomb (118938) *

      I propose that this is a task is best left suited to actual humans... ...and that I should be hired to assist.

      • by Locke2005 (849178)
        They're so cute when they're young and naive! Your first task is classifying the classic goatse, tubgirl, and lemon party pics, as well as the classic "2 girls 1 cup" video. The eye bleach is on the counter. Good luck!
        • by Adambomb (118938) *

          Proposition: The proportion of porn on the internet requiring eye bleach is but a tiny fraction of porn on the internet. A few instances of mind-rending pain would probably be worth that being ones occupation for the entire rest of the time!

  • So Galaxy Zoo doesn't need me anymore? That is the one activity where I was contributing to science to benefit all mankind.

    Oh well, I guess I'll go back to trying to beat Mario 64 or something equally pointless....

    • Re: (Score:3, Interesting)

      by $RANDOMLUSER (804576)
      I'm a Galaxy Zoomate, too, but I sure wouldn't mind an AI that would weed out even 75% of the boring eliptical galaxies, and let us concentrate on the pretty spirals and irregulars.
  • by Anonymous Coward on Wednesday June 09, 2010 @04:42PM (#32515826)

    PRO:

    Using neural networks allows for graceful degradation when classifying galaxies by indicating to what degree it believes this galaxy is similar to other galaxies of this type (that it has been trained on). A threshold can be set so that if confidence falls below this threshold, the image is flagged for human intervention.

    CONS:

    Neural nets are largely black boxes. They use learned statistical relationships to classify images, but they're unable to provide an explanation as to why they made the decision that they did.

    • that is not a "CON". there is an explanation: statistically, that is the class that the image belongs to.
      consider this: infants have a lot of inborn reflexes. and trust me, they don't understand why they do the things they do anymore than you would understand why your foot jumps when the doctor pokes your knee (I don't know the name of that reflex).
      there is absolutely no difference between using a trained human brain and using a trained simulated neural net.

      • Your analogy is bogus. We're talking about learning. Reflexes are prewired, not learned.
        Decision trees and KBANN can provide explanation.

        (I don't say that those explanation are always that useful. Giving useful explanations isn't an easy task even for humans, and it's ambigous as well. )

  • What's the point? (Score:1, Redundant)

    by Grishnakh (216268)

    Maybe I'm missing something, but what exactly is the point of going to all this effort to classify far-away galaxies? I can understand astronomers wanting to examine closer galaxies and see how they work and interact and all, but surely all the galaxies that are close enough for us to be able to see that much detail have already been known for some time, and are classified, and studied in far more detail than just classifying what kind of galaxy they are (spiral, barred spiral, elliptical, etc.). What goo

    • by Daniel Dvorkin (106857) * on Wednesday June 09, 2010 @05:16PM (#32516256) Homepage Journal

      There are a couple of answers to your question. The first is the answer to the more general question, "Why study the universe at all?" and the answer is "Because it's there." We want to understand the processes by which the universe we see around us was formed, what it's like now (to the degree that "now" has any meaning on cosmological scales) and where it's going. It is an awe-inspiring place, and becomes more so the more we learn about it.

      The second, with respect to the study of the Milky Way, is that we learn a lot about our galaxy by studying other galaxies. We don't have a good vantage point for studying the Milky Way, for obvious reasons. Hell, it wasn't until quite recently that we even knew what shape it was (barred spiral vs. plain spiral.) With the enormous number of galaxies out there, many of them similar to our own, at a variety of viewing angles from Earth, we can get a much better idea of what's going on in our own neighborhood than we could by restricting our observations to the Milky Way alone.

      • by Grishnakh (216268)

        The first is the answer to the more general question, "Why study the universe at all?" and the answer is "Because it's there."

        I'm not disputing that, only pointing out that observation resources are limited and should be put to best use.

        With the enormous number of galaxies out there, many of them similar to our own, at a variety of viewing angles from Earth, we can get a much better idea of what's going on in our own neighborhood than we could by restricting our observations to the Milky Way alone.

        I underst

        • by flowwolf (1824892)
          How do we know what the best use for observation resources is? why are ones that are closer to us any better than the ones far away? How do you determine where the good discoveries will be ? This seems like a politician's approach to science.
          • Re: (Score:3, Insightful)

            by Grishnakh (216268)

            That's what I was asking, because on the surface it seemed to me to make more sense studying nearby galaxies only. However, some other helpful responders pointed out that far-away galaxies allow us to see farther back in time (essentially, what we see of the far-away galaxies is how they appeared billions of years now, not how they appear now), and see how galaxies form and collide, and this might lead to insight into how our galaxy came into being.

            The difference between my questioning and the politicians

    • by $RANDOMLUSER (804576) on Wednesday June 09, 2010 @05:22PM (#32516346)
      Because the farther away they are, the farther back in time we're looking. By collecting images of galaxies at different stages of evolution (and different types of collisions) cosmologists are able to form a much better picture of how galaxies (and the universe in general) form and evolve.
      • by Grishnakh (216268)

        Ah, ok. This makes sense. Thanks!

        • further on, once we have a good idea about what happened in the universe, we can start checking that against the various sets of "laws of the universe" that we can generate, and decide which are valid (i.e. "is string theory ok, or do we need something else?").
          Afterwards, we can try to use the "correct" model of the universe to generate cool stuff (like quantum physics was used to properly describe semiconductors, and we got miniaturized electronics).

          I wasn't trying to exagerate. Usually in science, each ge

    • Re: (Score:2, Insightful)

      by dissy (172727)

      Pictures showing galaxies that are billions of light-years away make nice posters, but it seems totally pointless to put too much effort into these things, when there's so much we don't know about the stuff inside our own galaxy.

      But to learn about the stuff inside our galaxy, and how it came to be, we need to see how it looked in the past.
      Since you haven't gotten around to making that time machine yet, we can't do it that way ;}

      Instead we look at light from galaxies that have been traveling in space for an amount of time equal to how far back in time we want to see, and we discover such things as galaxy formation.

      This is ONLY possible to do by looking at distant and thus older galaxies. And it does teach us more about our own.

  • finally? (Score:3, Interesting)

    by Black Parrot (19622) on Wednesday June 09, 2010 @04:53PM (#32515952)

    I'm surprised they're just now getting around to this. It's a straightforward pattern classification problem, and there is a huge set of training examples to be used for training a neural network or other Learning Classifier System technologies.

  • Artificial Agent Aids Astral Analysis

  • This makes me very happy on one level and very sad on another.

    At the amateur end, the advances in technology have meant that what use to be done by a professional with mind blowingly expensive equipment or what was not at all possible because it hadn't been invented can now be done by a dedicated amateur with a reasonable but largish hobby budget. For the amount of money some spend on recreational vehicles and holiday homes an amateur can now do spectroscopy, deep imaging, even adaptive optics. It's not ope

  • Sorry to burst the bubble, but automatic classification of galaxies from sky survey data using machine learning techniques was accomplished in the early '90s by the SKICAT system developed at JPL and Caltech. http://adsabs.harvard.edu//abs/1995PASP..107.1243W [harvard.edu] is a good overview of the system and its accomplishments as of 1995.

"Life is a garment we continuously alter, but which never seems to fit." -- David McCord

Working...