An unrelated question I was pondering -- do any thumb-texting predictive text devices use predictive method more advanced than Markov chain? (essentially just looking at list of words that start with already typed prefix) That's how it seems to work on all the cell phones I tried. There's enough research out there to make a more advanced system. For instance, a word level Markov chain that offer words that are likely to follow first. Or an approach that would allow better generalization, such as decision tree predictor.
The fact that blisters are observed after another doesn't make it causation. Like you pointed out earlier, they could be both caused by a third, independent cause. The reason we assume Sun causes, and not merely corellates with skin damage is because of our prior knowledge, and subjective interpretation of data.
Here's another example -- you observe that every time a fox enters the chicken shed, you find some slain chickens the next day. Now, the observation is merely a corellation. Perhaps fox is just attracted to noise made by some unseen animal slaying the chickens. Despite that, the farmer will probably conclude fox did slay the chickens.
It's true that there could be 3rd cause, or even the causality goes the other way. But that's true of almost everything. We don't even know if the causality is local. For instance, here are some experiments which suggest that events in the future can cause events in the past -- http://citebase.eprints.org/cgi-bin/citations?id=o ai:arXiv.org:quant-ph/9801061
The question here is not whether they prove causality (because you can't), but how plausible their hypothesis is.
You can't prove causation in an indirect experiment, however, you'd be reasonable sure that Sun causes sunburn after experiencing the blisters. In real life we make causal inferences all the time without being able to "prove" them, and science does too.
Here's what Stephen Hawking had to say about it:
"Any physical theory is always provisional, in the sense that it is only a hypothesis; you can never prove it. No matter how many times the results of experiments agree with some theory, you can never be sure that the next time the result will not contradict the theory. On the other hand, you can disprove a theory by finding even a single observation that disagrees with the predictions of the theory"
So the relevant question is not whether they prove causation, but whether they support it. Whether or not observation supports a theory is a subjective judgement. Seeing the people sunburned after being in the Sun doesn't prove that Sun causes sunburn, but it supports the notion.
How do we know that Piraha people even understood what was required of them?
When people are asked to imitate, they tend to focus only on the important parts. For instance, if I wave, and ask a person to imitate me, they'll probably focus on my waving, and ignore my saccadic eye movements. One could then conclude that lack of short word for "saccadic eye movement" causes people to not see it.
The only way to reliably make sure they understand the task is to evaluate their performance on a validation set. IE one would give several examples of researcher tapping N times (>2), and assistant repeating N times. After the subject could successfully imitate the assistant on those examples, he should get previously unheard number of taps, and be asked to imitate it.
Yaroslav Artificial Intelligence in Python: yaroslav.hopto.org/pubwiki/index.php/ai-python
Programmers are low-level in terms of pyramid of responsibilities. There are lots of programmers, fewer managers and even fewer CEO's. When India's programmers can do as good of a job as American ones for fraction of the cost, it makes sense to export those tasks abroad.
Diamons not scarce?
on
The Diamond Age
·
· Score: 2, Interesting
"In its long history, De Beers has... and contended with Australian, Siberian, and Canadian diamond discoveries."
If DeBeers really inflates the price significantly, what is stopping those Australians Canadians and Russians from selling at lower price?
One of the concerns of the article was that Google would skew results towards article material. However almost all original research is published in forms of articles in peer reviewed journals, so having google result skew to articles is not necessarily a bad thing.
Magnetism affects all matter, not only ferrous materials.
Besides ferromagnetism, the strongest type and only manifested in certain materials/temperatures there's paramagnetism -- a weak attractive force and diamagnetism -- a weak repelling force.
While those types of magnetism are quite weak, they become noticeable in strong magnetic fields, and may cause objects to repel or attract.
We can't eliminate the forces caused by diamagnetism or paramagnetism, so we have to rely on our knowledge of the material to estimate what they are in that particular case. Anything left over after substracting those estimates must be a new force IF the estimates are correct.
I politely disagree. What they are doing is called "sterile-male technique." It takes advantage of the fact that many pests reproduce only once in a lifetime, so after female mated with sterile male, it will mate no more and hence produce no offspring. This technique was used in 1958 to effectively eradicate screwworm population in Southeast. The only thing different in this case is that instead of using radioactivity to sterilize large amounts of male insects, they are using genetic engineering. So if the technique is administered correctly, it may have a very strong effect on moth population.
Re:School isn't just to get a job
on
CS vs CIS
·
· Score: 1
Slashdot Mirror
An unrelated question I was pondering -- do any thumb-texting predictive text devices use predictive method more advanced than Markov chain? (essentially just looking at list of words that start with already typed prefix) That's how it seems to work on all the cell phones I tried. There's enough research out there to make a more advanced system. For instance, a word level Markov chain that offer words that are likely to follow first. Or an approach that would allow better generalization, such as decision tree predictor.
The fact that blisters are observed after another doesn't make it causation. Like you pointed out earlier, they could be both caused by a third, independent cause. The reason we assume Sun causes, and not merely corellates with skin damage is because of our prior knowledge, and subjective interpretation of data.
o ai:arXiv.org:quant-ph/9801061
Here's another example -- you observe that every time a fox enters the chicken shed, you find some slain chickens the next day. Now, the observation is merely a corellation. Perhaps fox is just attracted to noise made by some unseen animal slaying the chickens. Despite that, the farmer will probably conclude fox did slay the chickens.
It's true that there could be 3rd cause, or even the causality goes the other way. But that's true of almost everything. We don't even know if the causality is local. For instance, here are some experiments which suggest that events in the future can cause events in the past -- http://citebase.eprints.org/cgi-bin/citations?id=
The question here is not whether they prove causality (because you can't), but how plausible their hypothesis is.
You can't prove causation in an indirect experiment, however, you'd be reasonable sure that Sun causes sunburn after experiencing the blisters. In real life we make causal inferences all the time without being able to "prove" them, and science does too.
t hon
Here's what Stephen Hawking had to say about it:
"Any physical theory is always provisional, in the sense that it is only a hypothesis; you can never prove it. No matter how many times the results of experiments agree with some theory, you can never be sure that the next time the result will not contradict the theory. On the other hand, you can disprove a theory by finding even a single observation that disagrees with the predictions of the theory"
So the relevant question is not whether they prove causation, but whether they support it. Whether or not observation supports a theory is a subjective judgement. Seeing the people sunburned after being in the Sun doesn't prove that Sun causes sunburn, but it supports the notion.
Yaroslav
Artificial Intelligence in Python http://yaroslav.hopto.org/pubwiki/index.php/ai-py
How do we know that Piraha people even understood what was required of them?
When people are asked to imitate, they tend to focus only on the important parts. For instance, if I wave, and ask a person to imitate me, they'll probably focus on my waving, and ignore my saccadic eye movements. One could then conclude that lack of short word for "saccadic eye movement" causes people to not see it.
The only way to reliably make sure they understand the task is to evaluate their performance on a validation set. IE one would give several examples of researcher tapping N times (>2), and assistant repeating N times. After the subject could successfully imitate the assistant on those examples, he should get previously unheard number of taps, and be asked to imitate it.
Yaroslav
Artificial Intelligence in Python: yaroslav.hopto.org/pubwiki/index.php/ai-python
If one gets a weak high to start with, it will take them longer to get addicted. The scheme is only being considered for non-addicts.
Programmers are low-level in terms of pyramid of responsibilities. There are lots of programmers, fewer managers and even fewer CEO's. When India's programmers can do as good of a job as American ones for fraction of the cost, it makes sense to export those tasks abroad.
"In its long history, De Beers has ... and contended with Australian, Siberian, and Canadian diamond discoveries."
If DeBeers really inflates the price significantly, what is stopping those Australians Canadians and Russians from selling at lower price?
One of the concerns of the article was that Google would skew results towards article material. However almost all original research is published in forms of articles in peer reviewed journals, so having google result skew to articles is not necessarily a bad thing.
Besides ferromagnetism, the strongest type and only manifested in certain materials/temperatures there's paramagnetism -- a weak attractive force and diamagnetism -- a weak repelling force.
While those types of magnetism are quite weak, they become noticeable in strong magnetic fields, and may cause objects to repel or attract.
We can't eliminate the forces caused by diamagnetism or paramagnetism, so we have to rely on our knowledge of the material to estimate what they are in that particular case. Anything left over after substracting those estimates must be a new force IF the estimates are correct.
I politely disagree. What they are doing is called "sterile-male technique." It takes advantage of the fact that many pests reproduce only once in a lifetime, so after female mated with sterile male, it will mate no more and hence produce no offspring. This technique was used in 1958 to effectively eradicate screwworm population in Southeast. The only thing different in this case is that instead of using radioactivity to sterilize large amounts of male insects, they are using genetic engineering. So if the technique is administered correctly, it may have a very strong effect on moth population.
Education is expensive, but so is ignorance.
That's not news, it's been available for a while. I've gotten a copy two weeks ago. Yaroslav