The Milky Way Project Talk

Is there any way of finding out the accuracy of our ability to recognize desired target objects?

  • UncleClover by UncleClover

    I'm a very fast learner, but only if there's some sort of feedback going on letting me know what I'm doing right and what I'm doing wrong. Once an image that we notated has been processed by a professional, is there any way for us to look it up and find out how accurate or inaccurate our notations were?

    Feedback wouldn't even need to necessarily be an extra step, just having access to your own prior notations and being able to compare them to the "final conclusions" of the professional analysis would be useful. I know I'm not a pro and that this is something I don't necessarily have to be great at or anything, but I would like to be providing the highest signal-to-noise ratio possible via my output. I'm a total geek when it comes to that sort of thing - some find such study dull & boring, but I revel in it. 😉

    Posted

  • PattyD by PattyD

    I understand your frustration. I try very hard to produce as many bubbles as I can find. I think the more mature areas are the hardest because we see the stars that were generated and very little of the dust and gas. What gives me pause is that all the examples have only a few large bubbles indicated and I generally find about a hundred, but if no one else goes to that level of detail, my sightings will never get the required confirmation of 2 other sightings. I don't know how to reconcile the lack of information that we get. On the one hand if I know that 13 people found the same bubbles that I did, then I would feel more comfortable. If my ratio of confirmed bubbles to unconfirmed bubbles is low, then I am off track. Sigh... it's hard to know.

    Posted

  • masla1 by masla1

    For me, I found that the best way to learn was to go through the tutorial (both video and written), then do some myself, then go back to the tutorial. One thing I've learned from Moon Zoo (another of the zooniverse projects I take part in), is not to look too closely, because you'll see bubbles everywhere that way. Find the obvious, don't fret, and move on. These images are breathtaking and it's a very neat honor to be able to see them!
    Take care all.

    Posted

  • reidums by reidums scientist

    Part of why we're doing this is because there are too many images/bubbles for pros to go through them all, so your classifications are as definitive as anyone's. We are creating our procedure currently to take what everyone classifies and reduce it to an accurate catalog, so your finds will get combined with other peoples and if multiple people find the same bubble it will be more likely to make the final catalog. We're also currently figuring out a way to give more weight to the input from people who take more care (adjusting the bubbles' ellipticities and completeness and so on). If you've gone through the tutorial and spent some time on the site, you're probably about as good as any of the pros at this point!

    Posted

  • JLConawayII by JLConawayII

    One thing that will greatly help learning is to use these forums more. There should be multiple posts per day from many people asking all kinds of questions. Post pictures, ask "what is this wierd thing here?" or "how should I classify this?" Even if it's an image you've already classified, if you have questions or doubts post them. That's one thing that really caught me by surprise when I started analyzing for this project, the level of activity (or lack thereof) on this forum. I imagine there are plenty of discussions attached to individual images, but that can make it harder to track relevant information if it's spread out all over the place. I recommend centralizing it all here to make it easier for people to gain the knowledge they need to be proficient.

    Posted