Oops. Thanks for pointing that out. I had to read to the very end of the artile before that's made clear - something I should probably to more often in the future.
Like you, I will remain skeptical until I it is actually legal.
That is quite a striking difference. I have a couple of questions though, maybe for someone who can translate Chinese.
Firstly, what happens if I search for Tiananmen in Google China type in using Chinese characters rather than English?
Secondly, what's up with all the stuff at the bottom of the search results. i.e., [...pictograms....] US Digital Millennium Copyright Act [...pictograms...] ChillingEffects.org [...pictograms...] DMCA...
(Stupid lameness filter, won't let me paste in the Chinese characters. Or is Slashdot censoring posts from the Chinese?)
Visual Basic: You're a fifteen-year-old girl with her very own computer in her room, pinging random strangers on AIM and claiming to be a 23-year-old girl who wants to cyber with them. However, your efforts fail at convincing people, mostly because you aren't very imaginative and most of the things you're promising them are ideas you ripped off from other sources and changed slightly, leaving them less believable.
Humans do not solve the "pouring" problem analytically any more than they solve the "gravity" problem analytically. We use feedback.
When you see a ball thrown up your mind does not estimate the initial velocity of the ball, current wind direction and viscosity and use these to solve a differential equation to determine exactly where to place your hands to catch it. Instead, your eyes track the ball as it moves and you continually adjust you location and the position of your hands until you catch it.
Similarly, when pouring water I'm not doing fluid mechanics. I'm simply watching for when the water level in the receiving cup is close to full and then easing off the angle of the pouring cup until I'm done.
Do not confuse the abstract concepts we use to derive closed-form solutions to a problem and the actual mechanisms we employ to solve them in practice.
I agree with you, "decimal number plane" is a fairly idiosyncratic term to intriduce without any preliminary discussion but I think it has some merit if you take it to refer to R^2 but with a specific representation - i.e., a decimal one.
The point (1.1, 0.1) in the "binary number plane" would be (1.5, 0.5) in the "decimal number plane". By talking about the "decimal number plane" you build in the representation to the name.
I studied mathematics with Dr. Wildberger at the University of New South Wales (not South Wales University).
His classes were well thought out and engagingly presented. Although the link to his book is slashdotted at the moment, I'm sure that his take on trigonometry is both elegant and interesting.
While OS-X is _VERY_ nice any its pretty i'm not so sure about the keyboard usability. I can use a windows box without a mouse, i can't do that on a mac, maybe thats because i'm just not smart enough to know all the hotkeys.
Try a Mac with Quicksilver installed. It is amazing. If you love Ctrl-Space completion in programming IDEs such as Eclipse or cannot live without tab completion in a shell you'll be blown away with how Quicksilver changes the way you use your Mac.
For example, to open up slashdot in my browser (without the browser currently running) I type `Ctrl-Space sla `. To find my friend Joe's phone number in my address book (once again without the address book app running) I type `Ctrl-Space joe`. Joe's picture pops up and I hit right arrow to display his details. Want to find a document called "Meeting Notes.txt" that lives somewhere on your machine. Type `Ctrl-Space mee` and you'll see it and a list of other files starting with "mee". Hit `tab` and you can choose to open it, print it, trash it, move it, copy it, etc all with a couple more keystrokes.
No mouse required.
I had pretty much the same opinion of Macs as you do before I actually sat down and used one for a while. I think most of my bias came from using pre-OS X machines. With OS 9 and before I found it incredibly difficult to do anything on a Mac without constantly reaching for the mouse.
With OS X + Quicksilver I honestly barely ever touch the mouse and can do much more and more quickly than I have ever been able to on a Windows machine.
Oh, and by the way the word corporate comes from the word cooperation. And so, sense there must be more than one to have cooperation, it is appropriate to use they when talking about a corporation.
If you yourself are going to quote etymology to make a point against pedantry make sure you have your facts right first. The word corporation has nothing to do with the word cooperation apart from sharing many of its letters.
From the Merriam-Webster dictionary:
Main Entry: corporation
Pronunciation: "kor-p&-'rA-sh&n
Function: noun
Etymology: Late Latin corporatio, from Latin corporare to form into a body, from corpor- corpus body
So corporation is derived from the same root as "corpse" - it makes no claims about how many people need to be in it.
Please read Ludwig Wittgenstein about the use of language.
Wow! You're telling people to read about the philosophy of language after a spiel about people being "now-it-alls" for having "pedantic attention to grammer".
Real audiophiles (no that does not mean they have sex with their hi-fi) use analog because they claim that CD's loose to much of the music. Just because we don't "hear" it doesn't mean we don't "hear" it.
-1 Offtopic
Audiophiles listen to their sound systems, not music.
When the grandparent says crime is "low enough,"...
When was the last time you heard a grandparent say crime was "low enough"?
Now listen to your pappy, Freddy Jr., and consider yourself lucky. When I were a lad I was murdered three times on the way to school each day. In the afternoons we'd set the classroom on fire before kidnapping the teacher. After that there wasn't anything to do but rob the local bank a few times and spend the proceeds pumping H up our arms.
I'm also writing up my PhD, which also has a lot of equations in it. I knew about LaTeX from my undergrad days in mathematics and was happily writing my PhD in LaTeX until I discovered LyX.
According to the website LyX is a WYSIWYM document processor - What You See Is What You Mean. Basically, it's a Word-like frontend for LaTeX documents. Editing normal text is just like any word processor: ctrl-b for toggling bold, etc. But hit ctrl-m and you go into "math-mode" where standard LaTeX markup is understood and converted as you type into integral signs, sigmas, fractions, etc. Bibliographies, multifile documents, references and tables are all very well handled. It also has hooks for inlining XFig and other graphics including Grace plots.
At the end of the day (or paper), you still get beautiful LaTeX output but without having to worry nearly as much about compile errors. Why? Because, under the hood and behind the GUI you're editing LaTeX.
I'd just like to add some weight to the parent's claim that Foobar is a real gem. I wholeheartedly agree.
Foobar 2000 is now at a 0.71 release and is by far the most full featured and elegant media player I've seen on Windows. First time users might it a little sparse as the interface is very bland and not skinnable like Winamp or Sonique. However, under the hood is an amazing plugin architecture that is very well designed.
Some highlights for me:
Oggs, Mp3s, MODs, FLACs, AACs, and heaps more are all playable.
Keypresses for every single action the player can perform are configurable either as local or global (ie, you can control F2K while in another application)
The masstagger. This is amazing for reorganising your audio collection. It can guess album titles, song titles and artists from filenames, add arbitrary tags and store all this information is AWE2 within each file.
The album list. Once you've tagged all your files, you can heavily customise the way your files are organised and displayed in the album list.
The search facilities. These are second to none. Very fast and very powerful.
If you are using Windows and are listening to mp3s or oggs through WinAmp or Media Play I thoroughly recommend giving Foobar2000 a go.
I don't think Perl is hard to follow. I think Perl makes it extremely easy to write hard to follow code though. But hard to follow code can be written in any language.
There are good programmers and bad programmers. Good programmers can write clear, easy to follow code in most languages (exceptions being Malbolge and Intercal). Bad programmers manage to make life incredibly difficult no matter what their chosen tongue.
I'd be reluctant to use Perl at work for any code that has to be maintained by anyone except myself. It's a very expressive language and it's way to easy for different, equally competent coders to come up with incompatible idioms. In constrast, a language like Java places strong restrictions on the way you approach coding in it. On the one hand this is a good thing because you can quickly figure out what another coder is doing. It's not so good because it sometimes prevents you finding the neatest (and easiest to follow) way of doing something.
That said, I really enjoy coding in Perl for fun for the same reasons I like Go and composing music: intellectual stimulation. I usually write Perl programs to solve real problems but non-critical ones. That way I can have fun exploring different ways of solving problems.
And computing is so fundamentally simple. Its a game of N-Dimensional topology bounded by finite vectors in every dimension. There's no mystery involved. You just need to maintain a meta-model of the system and you can generate the rest.
I call bullshit.
Topology is a branch of mathematics, not a game.
How do vectors bound anything?
What the hell do you mean by "finite" vectors? It's almost a tautology. You said you're in an "N-dimensional topology" so the vectors are obviously of finite dimension. How are they going to be infinite?
Stringing together a whole bunch of terms you heard while dozing in your maths classes do not make you sound intelligent.
The global network of dreams is an interesting project for finding new music, books, films or websites.
Probably the most striking feature is its interface. You select a band, say Radiohead and a collection of related bands appear and slowly organize themselves so that similar bands are clustered and less similar ones seperated by what looks like some sort of simulated annealing process. Clicking on one of the names redraws the collection with the selected band at the centre.
First of all, it's not von Bayes. The guy was named Thomas Bayes.
Secondly, just because something is not state of the art does not mean it should be dismissed out of hand. You are right about Bayes classifiers making false assumptions about the independence of features but it has been suprisingly successful in practice, even when these assumptions have been violated. This paper shows that "...the accuracy of naive Bayes is not directly correlated with the degree of feature dependencies".
While kernel machines tend to be much more accurate (and quite cool theoretically), they are nowhere near as efficient (time and space-wise) to train. You want an intelligent spam filter to go easy on available resources and for this reason I don't think KMs are the way to go.
Another nice technology for text classification is Latent Semantic Analysis but once again, probably not the best tool for this particular job.
Slashdot Mirror
Did anyone else misread the URL for this new site?
Oops. Thanks for pointing that out. I had to read to the very end of the artile before that's made clear - something I should probably to more often in the future.
Like you, I will remain skeptical until I it is actually legal.
The ability to legally time-shift, format-shift and backup was made legal last month in Australia.
That is quite a striking difference. I have a couple of questions though, maybe for someone who can translate Chinese.
...
Firstly, what happens if I search for Tiananmen in Google China type in using Chinese characters rather than English?
Secondly, what's up with all the stuff at the bottom of the search results. i.e., [...pictograms....] US Digital Millennium Copyright Act [...pictograms...] ChillingEffects.org [...pictograms...] DMCA
(Stupid lameness filter, won't let me paste in the Chinese characters. Or is Slashdot censoring posts from the Chinese?)
e.g., newspaper is good for wrapping fish and chips
Humans do not solve the "pouring" problem analytically any more than they solve the "gravity" problem analytically. We use feedback.
When you see a ball thrown up your mind does not estimate the initial velocity of the ball, current wind direction and viscosity and use these to solve a differential equation to determine exactly where to place your hands to catch it. Instead, your eyes track the ball as it moves and you continually adjust you location and the position of your hands until you catch it.
Similarly, when pouring water I'm not doing fluid mechanics. I'm simply watching for when the water level in the receiving cup is close to full and then easing off the angle of the pouring cup until I'm done.
Do not confuse the abstract concepts we use to derive closed-form solutions to a problem and the actual mechanisms we employ to solve them in practice.
I agree with you, "decimal number plane" is a fairly idiosyncratic term to intriduce without any preliminary discussion but I think it has some merit if you take it to refer to R^2 but with a specific representation - i.e., a decimal one.
The point (1.1, 0.1) in the "binary number plane" would be (1.5, 0.5) in the "decimal number plane". By talking about the "decimal number plane" you build in the representation to the name.
I studied mathematics with Dr. Wildberger at the University of New South Wales (not South Wales University).
His classes were well thought out and engagingly presented. Although the link to his book is slashdotted at the moment, I'm sure that his take on trigonometry is both elegant and interesting.
Try a Mac with Quicksilver installed. It is amazing. If you love Ctrl-Space completion in programming IDEs such as Eclipse or cannot live without tab completion in a shell you'll be blown away with how Quicksilver changes the way you use your Mac.
For example, to open up slashdot in my browser (without the browser currently running) I type `Ctrl-Space sla `. To find my friend Joe's phone number in my address book (once again without the address book app running) I type `Ctrl-Space joe`. Joe's picture pops up and I hit right arrow to display his details. Want to find a document called "Meeting Notes.txt" that lives somewhere on your machine. Type `Ctrl-Space mee` and you'll see it and a list of other files starting with "mee". Hit `tab` and you can choose to open it, print it, trash it, move it, copy it, etc all with a couple more keystrokes.
No mouse required.
I had pretty much the same opinion of Macs as you do before I actually sat down and used one for a while. I think most of my bias came from using pre-OS X machines. With OS 9 and before I found it incredibly difficult to do anything on a Mac without constantly reaching for the mouse.
With OS X + Quicksilver I honestly barely ever touch the mouse and can do much more and more quickly than I have ever been able to on a Windows machine.
The parent said:
If you yourself are going to quote etymology to make a point against pedantry make sure you have your facts right first. The word corporation has nothing to do with the word cooperation apart from sharing many of its letters.
From the Merriam-Webster dictionary:
So corporation is derived from the same root as "corpse" - it makes no claims about how many people need to be in it.
Wow! You're telling people to read about the philosophy of language after a spiel about people being "now-it-alls" for having "pedantic attention to grammer".I'd start with a dictionary...
-1 Offtopic
Audiophiles listen to their sound systems, not music.
I heard a great story regarding inifinity and kids. Can't vouch for its veracity but here goes...
A mathematician is talking to a boy of about five years old and asks him, "What is the biggest number in the world?".
The boy replies, "540!".
The mathematician then asks, "What do you get if you add one to that?".
The boy furrows his brow for a moment and then his eyes open wide in astonishment.
Smiling, the mathematician thinks the boy has realised that no matter what number you think is the biggest, you can always add one to it.
"Wow!", the boy exclaims, "I was pretty close then!"
When was the last time you heard a grandparent say crime was "low enough"?
I read the above as Tolkien Ring revival! w00t!
I'm also writing up my PhD, which also has a lot of equations in it. I knew about LaTeX from my undergrad days in mathematics and was happily writing my PhD in LaTeX until I discovered LyX.
According to the website LyX is a WYSIWYM document processor - What You See Is What You Mean. Basically, it's a Word-like frontend for LaTeX documents. Editing normal text is just like any word processor: ctrl-b for toggling bold, etc. But hit ctrl-m and you go into "math-mode" where standard LaTeX markup is understood and converted as you type into integral signs, sigmas, fractions, etc. Bibliographies, multifile documents, references and tables are all very well handled. It also has hooks for inlining XFig and other graphics including Grace plots.
At the end of the day (or paper), you still get beautiful LaTeX output but without having to worry nearly as much about compile errors. Why? Because, under the hood and behind the GUI you're editing LaTeX.
If Java were open sourced, Sun would still....
Verbing weirds language.
While the situation is different, it may pay to remember Steve Fossett, an American who in 1998 crashed his hot air balloon off the cost of Australia after also attempt to fly around the world.
The rescue bill, which came to hundreds of thousands of dollars was footed by the Australian taxpayer.
It's not the first time we've helped out stricken adventurer's either.
I'm curious as to why the US or NZ bases just don't move him to an Australian base at Antartica and let him sort it out from there.
Foobar 2000 is now at a 0.71 release and is by far the most full featured and elegant media player I've seen on Windows. First time users might it a little sparse as the interface is very bland and not skinnable like Winamp or Sonique. However, under the hood is an amazing plugin architecture that is very well designed.
Some highlights for me:
If you are using Windows and are listening to mp3s or oggs through WinAmp or Media Play I thoroughly recommend giving Foobar2000 a go.
Kudos to the F2K team!
I don't think Perl is hard to follow. I think Perl makes it extremely easy to write hard to follow code though. But hard to follow code can be written in any language.
There are good programmers and bad programmers. Good programmers can write clear, easy to follow code in most languages (exceptions being Malbolge and Intercal). Bad programmers manage to make life incredibly difficult no matter what their chosen tongue.
I'd be reluctant to use Perl at work for any code that has to be maintained by anyone except myself. It's a very expressive language and it's way to easy for different, equally competent coders to come up with incompatible idioms. In constrast, a language like Java places strong restrictions on the way you approach coding in it. On the one hand this is a good thing because you can quickly figure out what another coder is doing. It's not so good because it sometimes prevents you finding the neatest (and easiest to follow) way of doing something.
That said, I really enjoy coding in Perl for fun for the same reasons I like Go and composing music: intellectual stimulation. I usually write Perl programs to solve real problems but non-critical ones. That way I can have fun exploring different ways of solving problems.
Thanks to Steve Kilbey of The Church for this one.
- Topology is a branch of mathematics, not a game.
- How do vectors bound anything?
- What the hell do you mean by "finite" vectors? It's almost a tautology. You said you're in an "N-dimensional topology" so the vectors are obviously of finite dimension. How are they going to be infinite?
Stringing together a whole bunch of terms you heard while dozing in your maths classes do not make you sound intelligent.Probably the most striking feature is its interface. You select a band, say Radiohead and a collection of related bands appear and slowly organize themselves so that similar bands are clustered and less similar ones seperated by what looks like some sort of simulated annealing process. Clicking on one of the names redraws the collection with the selected band at the centre.
Secondly, just because something is not state of the art does not mean it should be dismissed out of hand. You are right about Bayes classifiers making false assumptions about the independence of features but it has been suprisingly successful in practice, even when these assumptions have been violated. This paper shows that "...the accuracy of naive Bayes is not directly correlated with the degree of feature dependencies".
While kernel machines tend to be much more accurate (and quite cool theoretically), they are nowhere near as efficient (time and space-wise) to train. You want an intelligent spam filter to go easy on available resources and for this reason I don't think KMs are the way to go.
Another nice technology for text classification is Latent Semantic Analysis but once again, probably not the best tool for this particular job.