September 26, 2005
Help with Info Theory Maybe Earn some big(ger) bucks.
Paul Deignan was one of my earlier posts on "Newly Discovered Blogs" where I try to highlight new and interesting blogs. We have become friends over time and he has asked me to help him in designing a new way of looking at how people track info in what kinds of blogs. This will definately help bloggers understand their audiences better. And to sweeten the pot, Paul has decided to award 20 bucks $50 to any reader that clicks on his blog and is visitor number 15,000. If you leave a comment each time you visit if only to let him know where you came from (which blog you clicked on his from) he will also give a cash award to the referring URL.
So, click often (from here of course) on this entry. Info Theory and maybe you will be visitor number 15,000. I will put this link in all of my posts going back to September 1 so if you have a favorite, click from that post. Good Luck, and remember, this is for the advancement of science. (and the chance to earn 20 50 bucks.)
Twenty bucks! That's nothing compared to this site's prizes of having a barbeque and going bowling with G.M., boxes of rubber bands, and inspirational tapes. Maybe Info Theory could step up to the plate and offer World Series tickets. It would reach 15,000 before the end of the night.
Posted by Woody at September 27, 2005 06:48 PM
I heard a rumor that, if I commented here, I could win a new set of luggage. C'mon GM... I'm givin' away Suzuki Samarais over on my blog. Up the ante, my long-pocketed paisano.
Don't mess with Texas!
Posted by Nickie Goomba at September 27, 2005 11:04 PM
How about an autographed 1st edition copy of my dissertation in pdf (TBP)?
Posted by Paul Deignan at September 28, 2005 08:48 PM
Paul, I'm still trying to figure out the what you mean by the distribution. Can you give me an example or a hint? Also, how do you give an autographed copy of the dissertation if it is sent as a pdf file? I have an idea that I can understand quantum physics much easier than this.
Posted by Woody at September 28, 2005 08:58 PM
Hi Woody,
While this is not a good guess, one example would be:
1/3000 uniform
In other words, the 3000 hits between 13000 and 16000 are uniformly distributed. Here the parameter is redundant, but there are other distributions that have more degrees of freedom (normal, poisson, etc.) where one parameter would not be fixed by the interval alone.
Posted by Paul Deignan at September 29, 2005 12:35 PM
I should also verify this for the GM's Corner readers.
What GM is doing, for the benefit of his readers (he is a generous guy) is give $30 of the referring URL winner's proceeds directly to the reader, making the reader's prize $50.
There is a little game theory involved with this seconf phase of the experiment, so that proposal is not only generous, it is intelligent also.
Best of luck (successful estimation) to the GM readers!
Posted by Paul Deignan at September 29, 2005 01:57 PM
