I spent a great three days at Camp Hogg discussing Space Warps ( http://spacewarps.org ) with David, Dan FM and Ross: I talked them through how we analyze the database of classifications made by the 28,000 or so citizen scientists taking part in the project, in order to get feedback on the sense of what we were doing: the discussion was very useful, especially with regard to the supervised/unsupervised distinction that Hogg blogged about the next day. I hope that his group will step up with an unsupervised ML approach to the same data we are looking at in SW - that would make a very interesting comparison study, and potentially automate and speed up the lens finding process.
However, because I had not yet discussed the Space Warps analysis pipeline (SWAP) with the whole of my 8000-member collaboration, I told Camp Hogg that I could only show them our work in private! They completely understood - but then later, Hogg asked, can I blog about Space Warps, and I instinctively said yes, and he started a blog post as soon as I said yes, and all before we quickly realized that we had agreed to keep our discussion private. So I asked David would he mind cancelling his post, and he said, but I've started now - and so the post above is a compromise: Hogg got to make jokes about me censoring him, and I got to have my picture taken looking annoyed as I censured Hogg. My visits to NYU often involve shenanigans like this. It's why I like going there so much, I think.
Anyway, here's the blog post that should have been put up *before* I visited NYU:
This needs an explanation!
ReplyDeleteBTW photos and figures would be a great addition to Hogg's Research, but it would possibly make it more work.
NSA redacted?
ReplyDeleteAn explanation is indeed required! Here it is:
ReplyDeleteI spent a great three days at Camp Hogg discussing Space Warps ( http://spacewarps.org ) with David, Dan FM and Ross: I talked them through how we analyze the database of classifications made by the 28,000 or so citizen scientists taking part in the project, in order to get feedback on the sense of what we were doing: the discussion was very useful, especially with regard to the supervised/unsupervised distinction that Hogg blogged about the next day. I hope that his group will step up with an unsupervised ML approach to the same data we are looking at in SW - that would make a very interesting comparison study, and potentially automate and speed up the lens finding process.
However, because I had not yet discussed the Space Warps analysis pipeline (SWAP) with the whole of my 8000-member collaboration, I told Camp Hogg that I could only show them our work in private! They completely understood - but then later, Hogg asked, can I blog about Space Warps, and I instinctively said yes, and he started a blog post as soon as I said yes, and all before we quickly realized that we had agreed to keep our discussion private. So I asked David would he mind cancelling his post, and he said, but I've started now - and so the post above is a compromise: Hogg got to make jokes about me censoring him, and I got to have my picture taken looking annoyed as I censured Hogg. My visits to NYU often involve shenanigans like this. It's why I like going there so much, I think.
Anyway, here's the blog post that should have been put up *before* I visited NYU:
http://blog.spacewarps.org/2013/06/18/what-happens-to-your-markers-a-look-inside-the-space-warps-analysis-pipeline/
And now, finally, Hogg, you may comment! Thanks for waiting :-)
@Phil: We love you here at CampHogg!
ReplyDelete