Ah...but if spiritual forces always acted perpendicularly to the motion, like magnetic fields, they could act on the physical world without violating energy conservation.
It could be that words are matched to objects in the non-material spirit realm of the soul and that the part of the brain highlighted in this study is merely where those results are communicated back to the physical world. Or are you one of those un-American communist types who doesn't believe in souls?
...but for anyone with the patience to read the very short and readable one paragraph patent abstract it's pretty clear that the invention patented bears little relation to the description in the/. story. The/. story is, quite simply, a lie. But it's a lie about Microsoft, so I guess that's OK.
Comparing Firefly/Serenity to the South in the Civil war is one of the more whacked out things I've ever heard.
You have to be blind to fail to see the comparison, even down to the design of the clothing, the accents and the music.
A show where the entire premise is, arguably, about freedom?
So you think the people fighting for the South thought they weren't fighting for freedom? Do you imagine that when some redneck sticks a Confederate flag onto the side of his pickup it's saying anythig other than "I'm free to stick whatever the fuck I like onto my truck"?
The world isn't this simple black and white place. It's prefectly legitimate for writers to draw on the romanticism of the antebellum South without themselves becoming...oooohhh...completely evil.
How exactly do you determine how old a person is from film?
You use motion estimation to compute motion for every pixel in the frame. Using standard image segmentation algorithms you pull out the motions corresponding to pixels in each person in the camera view. You then track these for each segment over time. Now you switch to the frequency domain by computing the FFT of this data. People of different ages tend to produce peaks at different frequencies (this is kind of intuitively obvious). You compare against your training set and use that to judge the most likely age attached to each segment. You don't have to be 100% accurate - you're not trying to predict precisely who is a criminal but trying to increase the probability that any person you do check is one.
Because the original poster asked "But why would someone go out of their way to continue to use it?" It may be that he was just a generally all-round weird guy.
I guess that's all I needed to hear. If your statistical inferences don't match your logical inferences for the subset of problems that can be tackled by both then your ability to make inferences is seriously compromised. I suggest signing up for some philosophy courses on inference.
The actual spontaneous remission rate for this type of cancer (malignant melanoma) is 1/400
Good work. Now I can start computing. n=17, p = 1/400. (Probably
zero probability is within the margin of error
It most certainly is not. If the probability of spontaneous remission were zero then it'd be impossible to get 2 out of 17 spontaneous remissions so we'd be 100% sure that the two remissions were not spontaneous.
After we sample 1000 random forum posts across the Internet, we will be able to confirm the hypothesis with 99.9% confidence or more.
You're saying something completely bizarre. After you've collected 1000 sample posts you'll either catch my posts or not. If you do, then you'll be able to reject the hypothesis as clearly false. If you don't, then you certainly have little confirmation of the hypothesis. Have I understood you correctly? Are you claiming that after trawling 1000 posts, and finding that none of them have "Reinstate Pluto" in the sig, that this is good evidence that no posts at all have this in the sig?
Re:There's a gene that confers some resistance...
on
Humanity Gene Found?
·
· Score: 1
That sounds different again from the story I was discussing. Nonetheless, it's interesting stuff and it's a great example of how a gene can be human-specific while having little to do with 'humanity'.
What kind of ignoramus are you? Of course it wouldn't. With just 213 genes it'd qualify as slightly-better-than-average-human. But with 424 genes - now you're talking!
The problem is that in this case the correlation could have a thoroughly uninteresting casuality. For all we know this could a gene that makes you like bananas. And there really isn't a solid basis for saying this is a significant correlation. If you were to look at all of the non-intelligence-related biochemical processes taking place in mice, chimps and humans, and rate them by how much activity is taking place, any number of these could be correlated with this gene. I give a potential example in another post - it might be that the disease confers resistance to some virus and that the presence of this gene is nothing more than an indicator of how suitable each organism is as a host for this virus.
you are confusing mathematical logic and statistics.
No. But (1) I think you need the some skill in the former before you can embark on the latter and (2) in a suitable limit, statistics needs to match mathematical logic, and a good test of your statistical methodology is to examine it's behaviour in the limit and compare it with mathematical logic.
If X has a probability 0.00001 of happening spontaneously, say, and you observe X occuring twice out of 17 times, you have very good evidence that X is not occuring spontaneously. I'm not confusing significance level with margin of error. That's all you need. I'm stating that in this scenario (with hypothetical spontaneous remission rate) we have damn good evidence despite the small sample size. Good enough to publish. We don't have a particularly good estimate of the rate of remission caused by the new therapy. But we have excellent evidence that it is indeed causing remissions. If the rate of spontaneous remission is low enough, then this is enough to make publication worthwhile and the peer reviewers in Science were doing their job correctly.
Re:There's a gene that confers some resistance...
on
Humanity Gene Found?
·
· Score: 1
There is a tiny group in europe that has a mutation of a cell recptor the makes them getting HIV/AIDS more difficult.
I don't think this can be the same thing as what Steve Jones was discussing in the podcast. He was specifically interested in discussing the coevolution of humans and HIV and he gave the impression that the group of resistant humans wasn't all that small. It's a great podcast BTW.
You need sample sizes of 1000+ to get an accurate measure of how the population as a whole would vote. That's because people are trying to estimate a proportion of the entire population to within a few percent. But this result is quite different. As an extreme example: if someone claims that X is completely impossible then you only need to find a single instance of X to prove them wrong. Similarly, if they say the probability is less then 0.00001 and you find 2 examples in a sample size of 17, that is also completely acceptable and publication worthy. On the other hand, if you were testing the hypothesis that 50% of people vote for X with a sample size of 17, you're making a mistake.
Here, where the workers were testing whether or not a therapy works for a condition that is otherwise incurable and unlikely to simply go away, a sample of 17 might just do the trick. This is completely acceptable. Their sample size was certainly not too small unless the rate of spontaneous remission is high.
There's a gene that confers some resistance...
on
Humanity Gene Found?
·
· Score: 5, Interesting
...to HIV. Chimps have more of them than humans. It seems likely that SIV (simian immunodeficiency virus) has existed in chimps much longer than HIV in humans. As a result, Chimps with more copies of the gene have outlived their less well endowed relatives and now almost all chimps can coexist with SIV without showing symptoms of immunodeficiency. Apparently humans have started making similar adaptations and in some areas of the world there is now a generation of humans who seem to do a fairly good job of coexisting with HIV. But all humans still have many fewer of these genes than chimps.
But nobody would make the mistake of saying that this gene is the gene for 'chimpness'. It's just an accident of history that SIV arose before HIV.
I learned all of this from an excellent podcast whose name I dare not write for fear of offence...
But as far as I can tell from reading a few articles these patients were chosen because they had a close to zero chance of remission. If that is the case, 2/17 is significant. I've not seen the 'several people' you mention, although one person pointed out that remission rates using interferon are about 5%, suggesting that spontaneous rates are well below that. I've no idea where you get your "Positive outcome probability: p = 2/17" from. What matters is how likely 2 people out of 17 might recover spontaneously. The lower that probability, the more significant the 2 out of 17 becomes.
Suppose only 0.1% cancers of this type go into remission spontaneously. Then 2 out of 17 doing so is statistically significant because it's fairly unlikely. I've no idea what the spontaneous remission rates are, but neither do you.
With a simple confidence interval calculation...
You don't have the information required to make this computation. Without knowing spontaneous remission rates you don't have any kind of probability distribution to start working from. There is no "simple confidence interval computation". I think you're just blindly grabbing at any figures you can and fitting a completely meaningless normal distribution (or something) to then.
...Duke Nukem Forever was a geek joke. You know, like the coming of the Messiah, the second coming of Jesus or the arrival of the Maitreya Buddha. I didn't know there really was such a game in development. That's shattered my whole world view. I might have to rethink my views on religion now.
Slashdot Mirror
Ah...but if spiritual forces always acted perpendicularly to the motion, like magnetic fields, they could act on the physical world without violating energy conservation.
It could be that words are matched to objects in the non-material spirit realm of the soul and that the part of the brain highlighted in this study is merely where those results are communicated back to the physical world. Or are you one of those un-American communist types who doesn't believe in souls?
...but for anyone with the patience to read the very short and readable one paragraph patent abstract it's pretty clear that the invention patented bears little relation to the description in the /. story. The /. story is, quite simply, a lie. But it's a lie about Microsoft, so I guess that's OK.
The world isn't this simple black and white place. It's prefectly legitimate for writers to draw on the romanticism of the antebellum South without themselves becoming...oooohhh...completely evil.
Tell me, where in the world can you buy HD movies or watch HD TV programs that aren't compressed?
Version 1.0 conceived in 1968, implemented in 2006. There's home for Duke Nukem Forever then.
Because the original poster asked "But why would someone go out of their way to continue to use it?" It may be that he was just a generally all-round weird guy.
Just remember not to use the axe until you're close to the white door.
I guess not but I'll ask anyway...
...it's just that you need special glasses to see what's going on.
A million monkeys and we still don't see no Hamlet
That sounds different again from the story I was discussing. Nonetheless, it's interesting stuff and it's a great example of how a gene can be human-specific while having little to do with 'humanity'.
What kind of ignoramus are you? Of course it wouldn't. With just 213 genes it'd qualify as slightly-better-than-average-human. But with 424 genes - now you're talking!
The problem is that in this case the correlation could have a thoroughly uninteresting casuality. For all we know this could a gene that makes you like bananas. And there really isn't a solid basis for saying this is a significant correlation. If you were to look at all of the non-intelligence-related biochemical processes taking place in mice, chimps and humans, and rate them by how much activity is taking place, any number of these could be correlated with this gene. I give a potential example in another post - it might be that the disease confers resistance to some virus and that the presence of this gene is nothing more than an indicator of how suitable each organism is as a host for this virus.
If X has a probability 0.00001 of happening spontaneously, say, and you observe X occuring twice out of 17 times, you have very good evidence that X is not occuring spontaneously. I'm not confusing significance level with margin of error. That's all you need. I'm stating that in this scenario (with hypothetical spontaneous remission rate) we have damn good evidence despite the small sample size. Good enough to publish. We don't have a particularly good estimate of the rate of remission caused by the new therapy. But we have excellent evidence that it is indeed causing remissions. If the rate of spontaneous remission is low enough, then this is enough to make publication worthwhile and the peer reviewers in Science were doing their job correctly.
Here, where the workers were testing whether or not a therapy works for a condition that is otherwise incurable and unlikely to simply go away, a sample of 17 might just do the trick. This is completely acceptable. Their sample size was certainly not too small unless the rate of spontaneous remission is high.
But nobody would make the mistake of saying that this gene is the gene for 'chimpness'. It's just an accident of history that SIV arose before HIV.
I learned all of this from an excellent podcast whose name I dare not write for fear of offence...
But as far as I can tell from reading a few articles these patients were chosen because they had a close to zero chance of remission. If that is the case, 2/17 is significant. I've not seen the 'several people' you mention, although one person pointed out that remission rates using interferon are about 5%, suggesting that spontaneous rates are well below that. I've no idea where you get your "Positive outcome probability: p = 2/17" from. What matters is how likely 2 people out of 17 might recover spontaneously. The lower that probability, the more significant the 2 out of 17 becomes.
...Duke Nukem Forever was a geek joke. You know, like the coming of the Messiah, the second coming of Jesus or the arrival of the Maitreya Buddha. I didn't know there really was such a game in development. That's shattered my whole world view. I might have to rethink my views on religion now.