The Bayesian is not even shown being Bayesian. Bayesianism is assigning prior probabilities to outcomes and then updating them based on the data, which he isn't shown doing in the comic. You can read between the lines and assume he's assigning a very low prior probability to the event that the sun has actually gone nova, but it's not shown. The problem is, a frequentist can just as easily translate probabilities into odds ratios, and thus come up with good betting strategies. The Bayesian is not doing anything that a good frequentist can't. Which brings us to the next point.
The frequentist is, yes, being a frequentist, but he's also being a fucking idiot (i.e. a strawman). The really stupid part is that he uses the 5% criterion, which even in social sciences is usually only enough to conclude "hey, there might be something there, let's do a better study." Frequentists are allowed to (and do) require ``extraordinary evidence for extraordinary claims." For such an amazingly unlikely event as the sun going nova, a non-strawman frequentist would require something like 0.0000001% significance level, which the machine obviously cannot provide. The non-strawman frequentist wouldn't bother using the machine at all, since it obviously cannot satisfy his criteria (the best significance it can give is 1/36, which isn't good enough).
The significance level here is "the probability of observing the evidence we observed, if the null is true," in this case the null hypothesis being that the sun has not exploded. A frequentist is allowed to use "prior information" to set the significance level. He just isn't allowed to say something like "Before I look at the machine, I think there is a p% chance that the sun has exploded, and a (1-p)% chance that it has not." (This is a Bayesian approach which, as I mentioned, the Bayesian doesn't actually do in the comic.) A frequentist is only allowed to judge how likely it is that the evidence is a false positive, and this likelihood is the significance level.
I hope this is clear; it's hard to explain this stuff sometimes.
he seems to prefer knowing the time to the nearest 5 or 15 minutes or whatever, and this watch gives it to him without his needing to process it mentally. since he calls the process "trivial," i can only conclude that, yes, he could do it himself like anyone else on the planet, but why bother?
more likely, he just think it's cool and will get over it in a few days when he finds himself toggling to the standard clock mode for every appointment, but whatever.
Well, yeah, TeX isn't ideal for that, and it's not really meant to be. It could be adapted for it, or at least have a wrapper written, but I guess no one is interested in doing so.
PDF has some support for scalable graphics. Can you export to PDF and use that in the latex file?
It's basically equivalent to quoting a portion of a work for a book review, which is fair use. Google's profitability is irrelevant.
People should be thankful for having an opt-out robots.txt at all. It would in most cases not violate copyright for Google to ignore it completely (it might violate the site's TOS, but it is questionable whether even this holds any weight); they are just courteous enough to honor it. robots.txt is there mostly to prevent servers from being overloaded, or to keep content private, not to enforce copyright on publicly-facing content.
No one cares what you "find weak" or "struggle understanding." I'm sure there's quite a bit.
pdflatex can accept basically anything other than EPS. yeah, including graphics is still a major pain, but converting to EPS is not typically a problem.
I'm just saying, don't look to the postmodernist brigade to do much. They knew that the liberal arts are potentially threatening, and defanged them as much as possible. The current crop of liberal arts students are mostly either self-defeating navel-gazers; happy sellouts for the corporate-state party line; or, at best, cogs in the NGO machine.
You're not going to find any revolutionary ideas in this generation as it stands now.
This is the same as any NLP crowd-sourcing tool; it's simply designed with a focus on correlating vocabulary with prejudicial sentiment. No one is planning to use it to pass restrictive laws. It's just useful for people who are involved in a country, but are not fluent speakers of $foo or involved in the right subcultures, to know that a certain word has now acquired a negative connotation.
Maybe those people should butt out, sure, but you're jumping the gun a bit, here. If they could force everyone to use ``political correct speeches", they wouldn't need this app in the first place.
yes, until you need the services of a professional who belongs to a guild which fixes their prices high so that those professionals can afford all of those gadgets, comforts, cars, square feet, and travel.
Totally off-topic, but that comic completely misses the point of the frequentist vs. Bayesian kerfuffle. I do mostly Bayesian stats, so I don't have a problem with it, but christ, Randall really missed the ball on that one.
``Your grade is C. To improve your grade in the future, you need to do the following:
use 25-30 words per sentence; include more words from the wordnet entry for the topic of your essay; avoid simplistic or run-on sentences as measured by number of noun and verb phrases detected by our proprietary NLP tokenizer.
As a helpful reminder, our preparatory guides are available as a subscription service and include 100 practice submissions per week; only $29.95 per month."
Prior restraint. If a website entirely forbids the posting of an entire class of image, let's say abstract art, then because this law establishes a criminal charge for doing so, it should be considered prior restraint in that case.
Since this law does not rule out cases like the above (and it would be incredibly difficult or impossible to write it in such a way that it did), and there almost certainly are websites with such restrictions in their TOS, it thus violates the first amendment quite grievously. As long as national security isn't involved, courts take prior restraint quite seriously.
There is no way in hell that this clause will stand. They'll remove it soon and make hay over how it shows that they care about freedom, blah, blah, blah.
yeah, you're right. it means that they have a commission, rather than that they are commissioned, which is what i thought. thanks. sorry for the confusion.
In particular, note subchapter 10, article 133: ``Any commissioned officer, cadet, or midshipman who is convicted of conduct unbecoming an officer and a gentleman shall be punished as a court-martial may direct."
The veterans I know refer to this as the ``fuck you'' clause, as it clearly covers practically anything and only applies to the peons (ranking officers are noncommissioned, or `non-coms').
My pending patent has already been assigned to my superiors. I'm not sure exactly how this is implemented, but i think the point of the law is to allow this kind of thing (it's how capitalism is supposed to work, after all) while stopping the kind of patent troll which buys up a collection of older, tactically useful, patents for pennies and then spams lawsuits with them.
A non-practicing entity could still acquire patents and troll with them, if they were the original assignee. So, it's a conservative law; it attacks only the most flagrant case of trolling, while giving the benefit of doubt elsewhere. Upon a few seconds of reflection, one realizes that this is a good thing.
I find trigger warnings asinine. In their current form, they evolved in response to shorter and shorter fiction. For example, a full-length movie or novel will, generally, have some kind of review or advertisement about its content, so no problem; anyone can make an informed decision. Trigger warnings in their current form, afaict, caught on first with shitty four paragraph-long fanfic, so by the time you've read it, you've already been subjected to a crude depiction of pony/vampire rape written at a sixth-grade level. Hence, a prologue comprising a list of `triggers'.
Now we have gone further; the title of a talk, ``sex +/- drugs: known vulnerabilities and exploits," itself is enough to merit a warning. Maybe you've deduced my point already. If that title is enough to `trigger' someone, how isn't ``trigger warning: contains a very oblique reference to drug-related rape" capable of same? Maybe we instead can use a more gentle warning, perhaps, ``trigger warning: contains reference to chemicals and forced sex"? But the phrase ``forced sex" is a bit rough. Maybe instead we can say ``trigger warning: contains reference to chemicals, sexuality, and vulnerability," or ``trigger warnings: contains reference to drugs, sex, vulnerability and exploitation"?
Do you see what i did there? The title of the talk is already a trigger warning (albeit accidentally, as the content was not actually about rape). It just wasn't sanctioned as one by the self-appointed gatekeepers of culture and sanity, so out it goes.
This has been about this particular trigger warning which is especially ridiculous; however, I think they all are. As someone has said about this topic: Give me life, give me pain. Give me myself again. Life isn't going to go away; if you try to sanitize it, you pathologize yourself and everyone else.
Slashdot Mirror
The Bayesian is not even shown being Bayesian. Bayesianism is assigning prior probabilities to outcomes and then updating them based on the data, which he isn't shown doing in the comic. You can read between the lines and assume he's assigning a very low prior probability to the event that the sun has actually gone nova, but it's not shown. The problem is, a frequentist can just as easily translate probabilities into odds ratios, and thus come up with good betting strategies. The Bayesian is not doing anything that a good frequentist can't. Which brings us to the next point.
The frequentist is, yes, being a frequentist, but he's also being a fucking idiot (i.e. a strawman). The really stupid part is that he uses the 5% criterion, which even in social sciences is usually only enough to conclude "hey, there might be something there, let's do a better study." Frequentists are allowed to (and do) require ``extraordinary evidence for extraordinary claims." For such an amazingly unlikely event as the sun going nova, a non-strawman frequentist would require something like 0.0000001% significance level, which the machine obviously cannot provide. The non-strawman frequentist wouldn't bother using the machine at all, since it obviously cannot satisfy his criteria (the best significance it can give is 1/36, which isn't good enough).
The significance level here is "the probability of observing the evidence we observed, if the null is true," in this case the null hypothesis being that the sun has not exploded. A frequentist is allowed to use "prior information" to set the significance level. He just isn't allowed to say something like "Before I look at the machine, I think there is a p% chance that the sun has exploded, and a (1-p)% chance that it has not." (This is a Bayesian approach which, as I mentioned, the Bayesian doesn't actually do in the comic.) A frequentist is only allowed to judge how likely it is that the evidence is a false positive, and this likelihood is the significance level.
I hope this is clear; it's hard to explain this stuff sometimes.
i disagree.
he seems to prefer knowing the time to the nearest 5 or 15 minutes or whatever, and this watch gives it to him without his needing to process it mentally. since he calls the process "trivial," i can only conclude that, yes, he could do it himself like anyone else on the planet, but why bother?
more likely, he just think it's cool and will get over it in a few days when he finds himself toggling to the standard clock mode for every appointment, but whatever.
anyone is clever enough to do that approximating in his head. some people find different versions of presentation to be aesthetically pleasing.
Well, yeah, TeX isn't ideal for that, and it's not really meant to be. It could be adapted for it, or at least have a wrapper written, but I guess no one is interested in doing so.
PDF has some support for scalable graphics. Can you export to PDF and use that in the latex file?
Oh, wait, I just looked it up. Docbook doesn't have math elements! To include math, you can passthrough either MathML or ... wait for it ... LaTeX code.
Hilarious. You're an idiot.
It's basically equivalent to quoting a portion of a work for a book review, which is fair use. Google's profitability is irrelevant.
People should be thankful for having an opt-out robots.txt at all. It would in most cases not violate copyright for Google to ignore it completely (it might violate the site's TOS, but it is questionable whether even this holds any weight); they are just courteous enough to honor it. robots.txt is there mostly to prevent servers from being overloaded, or to keep content private, not to enforce copyright on publicly-facing content.
No one cares what you "find weak" or "struggle understanding." I'm sure there's quite a bit.
pdflatex can accept basically anything other than EPS. yeah, including graphics is still a major pain, but converting to EPS is not typically a problem.
structure is one thing, output is another. show me some docbook-rendered math that doesn't suck.
coprophagy.
cacophagy isn't typically considered a word, but read in greek it would mean "eating of evil".
Yes, yes, I said I agree with a lot of that!
I'm just saying, don't look to the postmodernist brigade to do much. They knew that the liberal arts are potentially threatening, and defanged them as much as possible. The current crop of liberal arts students are mostly either self-defeating navel-gazers; happy sellouts for the corporate-state party line; or, at best, cogs in the NGO machine.
You're not going to find any revolutionary ideas in this generation as it stands now.
Oh, calm down.
This is the same as any NLP crowd-sourcing tool; it's simply designed with a focus on correlating vocabulary with prejudicial sentiment. No one is planning to use it to pass restrictive laws. It's just useful for people who are involved in a country, but are not fluent speakers of $foo or involved in the right subcultures, to know that a certain word has now acquired a negative connotation.
Maybe those people should butt out, sure, but you're jumping the gun a bit, here. If they could force everyone to use ``political correct speeches", they wouldn't need this app in the first place.
yes, until you need the services of a professional who belongs to a guild which fixes their prices high so that those professionals can afford all of those gadgets, comforts, cars, square feet, and travel.
i'm mostly referring to doctors.
Maybe we could find such people among the liberal arts students currently having economic trouble.
I agree with your statement of the problem but you're not going to find the solution there...
You're being overly generous. Everything you describe after "Beyond that" is beyond the state-of-the-art in NLP to do consistently.
Totally off-topic, but that comic completely misses the point of the frequentist vs. Bayesian kerfuffle. I do mostly Bayesian stats, so I don't have a problem with it, but christ, Randall really missed the ball on that one.
Good point.
``Your grade is C. To improve your grade in the future, you need to do the following:
use 25-30 words per sentence; include more words from the wordnet entry for the topic of your essay; avoid simplistic or run-on sentences as measured by number of noun and verb phrases detected by our proprietary NLP tokenizer.
As a helpful reminder, our preparatory guides are available as a subscription service and include 100 practice submissions per week; only $29.95 per month."
Prior restraint. If a website entirely forbids the posting of an entire class of image, let's say abstract art, then because this law establishes a criminal charge for doing so, it should be considered prior restraint in that case.
Since this law does not rule out cases like the above (and it would be incredibly difficult or impossible to write it in such a way that it did), and there almost certainly are websites with such restrictions in their TOS, it thus violates the first amendment quite grievously. As long as national security isn't involved, courts take prior restraint quite seriously.
There is no way in hell that this clause will stand. They'll remove it soon and make hay over how it shows that they care about freedom, blah, blah, blah.
yeah, you're right. it means that they have a commission, rather than that they are commissioned, which is what i thought. thanks. sorry for the confusion.
Yes, it's called the Uniform Code of Military Justice.
In particular, note subchapter 10, article 133: ``Any commissioned officer, cadet, or midshipman who is convicted of conduct unbecoming an officer and a gentleman shall be punished as a court-martial may direct."
The veterans I know refer to this as the ``fuck you'' clause, as it clearly covers practically anything and only applies to the peons (ranking officers are noncommissioned, or `non-coms').
My pending patent has already been assigned to my superiors. I'm not sure exactly how this is implemented, but i think the point of the law is to allow this kind of thing (it's how capitalism is supposed to work, after all) while stopping the kind of patent troll which buys up a collection of older, tactically useful, patents for pennies and then spams lawsuits with them.
A non-practicing entity could still acquire patents and troll with them, if they were the original assignee. So, it's a conservative law; it attacks only the most flagrant case of trolling, while giving the benefit of doubt elsewhere. Upon a few seconds of reflection, one realizes that this is a good thing.
This doesn't seem to generalize beyond tawdry goals of the form "receive X tokens of appreciation from others."
What happens if I have a goal like "understand quantum mechanics"?
this problem has already been solved.
while slashdot is, of course, both slow and alarmist.
I find trigger warnings asinine. In their current form, they evolved in response to shorter and shorter fiction. For example, a full-length movie or novel will, generally, have some kind of review or advertisement about its content, so no problem; anyone can make an informed decision. Trigger warnings in their current form, afaict, caught on first with shitty four paragraph-long fanfic, so by the time you've read it, you've already been subjected to a crude depiction of pony/vampire rape written at a sixth-grade level. Hence, a prologue comprising a list of `triggers'.
Now we have gone further; the title of a talk, ``sex +/- drugs: known vulnerabilities and exploits," itself is enough to merit a warning. Maybe you've deduced my point already. If that title is enough to `trigger' someone, how isn't ``trigger warning: contains a very oblique reference to drug-related rape" capable of same? Maybe we instead can use a more gentle warning, perhaps, ``trigger warning: contains reference to chemicals and forced sex"? But the phrase ``forced sex" is a bit rough. Maybe instead we can say ``trigger warning: contains reference to chemicals, sexuality, and vulnerability," or ``trigger warnings: contains reference to drugs, sex, vulnerability and exploitation"?
Do you see what i did there? The title of the talk is already a trigger warning (albeit accidentally, as the content was not actually about rape). It just wasn't sanctioned as one by the self-appointed gatekeepers of culture and sanity, so out it goes.
This has been about this particular trigger warning which is especially ridiculous; however, I think they all are. As someone has said about this topic: Give me life, give me pain. Give me myself again. Life isn't going to go away; if you try to sanitize it, you pathologize yourself and everyone else.