Historically most military deaths have occurred in the base camps because of poor hygiene, but the Roman Army engineers didn't just build ballistas, they enforced strict camp procedures for latrines and water sources. Most people know about their discipline and weaponry, but hygiene was a major factor in their effectiveness. They could use smaller forces, with all of the logistical benefits that implies, and still field more effective soldiers than their opponents. They didn't know germ theory, they just worked from empirical evidence (pun intended).
What you've quoted describes an observational study, one in which you observe outcomes but can't or don't intervene. Correlation can be because one of the outcomes causes the other, but in order to prove causality you need to perform controlled experiments in which you can vary potential causal factors and observe the effects of your intervention. You cannot prove causality with observational studies because there is always the possibility of spurious correlation (things just happened to happen together) or of some unobserved factor being an underlying cause for both the phenomena that you think are linked. For example, in Finland the number of deaths per month by drowning and the monthly per-capita consumption of ice cream are very strongly correlated. Does that mean that eating ice cream increases the hazard of drowning? No, it means that Finns eat ice cream more frequently during the same months they go swimming, i.e., in the summer. The actual cause in both cases is the seasonality of both behaviors.
It's not revisionist. The term "hacker" has a history going back more than 50 years at MIT. Although its use has changed over the years, it definitely carried a connotation of using skill, imagination, and wits.
Okay, you're wrong. GP got it correct. On the last page of the penultimate chapter, Rico mumbles something. When asked what language it was, he responds "Tagalog. My native language."
As a fellow statistician, you're welcome. And thanks for correcting the misconceptions about null vs impossible events. I was trying to think of a reply for that thread that didn't get into measure theory when I saw your response and thought, "good, I don't have to do that one."
Yes, but a logistic regression is better suited to modeling a discrete choice outcome (like war or no war).
True, but I was trying to point out that some statistical models yield probability estimates, while others yield predictions. The original claim that "Statistics can never predict the outcome, they can only give you a probability of an outcome" is wrong, and sure as heck shouldn't be labeled as "insightful".
A linear regression can model a predicted value of a continuous variable, but that prediction will have a confidence interval associated with it. In other words, a linear regression will predict a range of outcomes with some predefined probability of occurring (by convention, 95% is usually used as the confidence interval with some exceptions according to application).
Nitpick time - A linear regression yields an expected outcome, which under certain distributional assumptions for the residuals can be used to produce range estimators such as confidence intervals, prediction intervals, or tolerance intervals. Which is more appropriate depends on the intended use of the model.
Statistics can never predict the outcome, they can only give you a probability of an outcome.
Sorry, but you don't know what you're talking about. There are many kinds of statistical models. Models such as logisitic regression map inputs to probabilities of outcomes, but models such as linear regression map the inputs to the actual predicted outcomes, not their probabilities.
A statistical model can be predictive without being causal, i.e., the inputs don't necesarrily cause the outcome, but they are observed to occur jointly. Hence the old saying "correlation is not the same thing as causality". There are lots of good examples of this, one of my favorites is that the number of deaths by drowning per month in Finland is highly correlated with the ice cream consumption per month. People don't drown due to the ice cream - the correlation is because the number of people drowning in a given month is proportional to how many people go swimming, and many fewer people go swimming or eat ice cream in Finland's winter months.
The full sentence was "It also marked the first time this year that an operating system security update from Apple did not patch a vulnerability disclosed by the January Month of Apple Bugs project." To quote Inigo, "I don't think that means what you think it means."
That's why I put the word "like" after "Quantrix/Improv". Here, I'll use it a few more times for you: Quantrix was very much like Improv in philosophy, and I would very much like to see a spreadsheet like either one of 'em. The fact that Sun has the OpenStep code-base for Quantrix in their archives gives me some hope that it could happen. Then again, I've always been an optimist.
Sun already owns the rights to Lighthouse Design's application suite. Since these were originally developed for NeXTstep/OpenStep, they should be relatively easy to migrate to Cocoa. I'd sure like to see an Improv/Quantrix like spreadsheet tool put a stake through the heart of Excel!
I disagree that Stallman has attempted to redefine the English language. He himself noted the ambiguity of the word "free", and clarified his intent with the (in)famous quote "Free as in freedom, not as in beer."
Okay, I call bullshit. You clearly don't know anything about MIT, and have an axe to grind when it comes to academics.
Fact 1: MIT has granted Full Professorships to people without degrees. They care about performance and ability more than about degrees.
Fact 2: They also care about integrity. A place like MIT earns and maintains its reputation based on both the quality and the integrity of the work done there. Integrity is where the dean screwed up, and why she is being canned.
Took about 2 minutes to find those stories and provide links.
Compare the time of your post and the time of mine. I heard this first on Slashdot, checked the.gov links, and did a search of my own. At the time I made my post I hadn't been able to find anything on the sites I mentioned, all of which I consider to be major news outlets.
Easier to believe it's a corporate media conspiracy eh? I could provide a few hundred more but you truthers aren't worth the time.
I don't know what you mean by "truthers" or why you think I am one, but it looks to me like you've got some kind of a grudge. I was merely commenting on something that was fact at the time of my post.
I'm fascinated that there's nothing about this on NY Times, CNN, or BBC.
Re:It would be a great first language
on
Beginning Ruby
·
· Score: 1
As an MIT alum, I'm a big fan of SICP. However, not many curricula use it and a lot of people "don't get it." I'm not sure why you addressed your response to my earlier message, I wasn't one of the advocates of "learn assembly and C first". I'm saying that Ruby wouldn't be a bad choice for a first language because it doesn't bog you down in lots of overhead early on, but has enough richness that you can carry on with it through some very sophisticated concepts.
Re:It would be a great first language
on
Beginning Ruby
·
· Score: 1
I think you're grabbing the wrong end of the question. From either an instructor or student's viewpoint the concern should be whether the language requires you to know all those features to do even simple programs. As you point out, Java requires a lot of overhead to write "Hello world". Ruby doesn't. The reason I think Ruby would be a good teaching language is that you can start small, but take it as far as you want with regard to OOP, functional programming, etc., all without having to switch languages along the way. Ruby keeps simple things simple, but the power is all there if you want to draw on it.
You can criticize Ruby for its speed in the various shootouts, but even there I hearken back to the old saying "There are no slow languages, only slow implementations." Any runtime slowness is usually offset by reductions in my development time - I find it easier to prototype an idea in Ruby, then translate to Java/C++ if I need more speed, and for my work Ruby will cough up the answer within a second or two of run time 95% of the time. However, I do academic research with a lot of one-off code - your mileage may vary.
Slashdot Mirror
The lower GB is because it's flash based instead of having a mini hard drive. That extends the battery life and gives them the 8mm profile.
They pinned it on a dead guy. If he were still alive to defend himself, he'd probably insist on a blood test.
Historically most military deaths have occurred in the base camps because of poor hygiene, but the Roman Army engineers didn't just build ballistas, they enforced strict camp procedures for latrines and water sources. Most people know about their discipline and weaponry, but hygiene was a major factor in their effectiveness. They could use smaller forces, with all of the logistical benefits that implies, and still field more effective soldiers than their opponents. They didn't know germ theory, they just worked from empirical evidence (pun intended).
sanitation and clean water supplies have arguably done more to extend the average lifespan than the entire field of medicine.
What you've quoted describes an observational study, one in which you observe outcomes but can't or don't intervene. Correlation can be because one of the outcomes causes the other, but in order to prove causality you need to perform controlled experiments in which you can vary potential causal factors and observe the effects of your intervention. You cannot prove causality with observational studies because there is always the possibility of spurious correlation (things just happened to happen together) or of some unobserved factor being an underlying cause for both the phenomena that you think are linked. For example, in Finland the number of deaths per month by drowning and the monthly per-capita consumption of ice cream are very strongly correlated. Does that mean that eating ice cream increases the hazard of drowning? No, it means that Finns eat ice cream more frequently during the same months they go swimming, i.e., in the summer. The actual cause in both cases is the seasonality of both behaviors.
It's not revisionist. The term "hacker" has a history going back more than 50 years at MIT. Although its use has changed over the years, it definitely carried a connotation of using skill, imagination, and wits.
As a fellow statistician, you're welcome. And thanks for correcting the misconceptions about null vs impossible events. I was trying to think of a reply for that thread that didn't get into measure theory when I saw your response and thought, "good, I don't have to do that one."
A statistical model can be predictive without being causal, i.e., the inputs don't necesarrily cause the outcome, but they are observed to occur jointly. Hence the old saying "correlation is not the same thing as causality". There are lots of good examples of this, one of my favorites is that the number of deaths by drowning per month in Finland is highly correlated with the ice cream consumption per month. People don't drown due to the ice cream - the correlation is because the number of people drowning in a given month is proportional to how many people go swimming, and many fewer people go swimming or eat ice cream in Finland's winter months.
The full sentence was "It also marked the first time this year that an operating system security update from Apple did not patch a vulnerability disclosed by the January Month of Apple Bugs project." To quote Inigo, "I don't think that means what you think it means."
...and that's what really matters, eh?
That's why I put the word "like" after "Quantrix/Improv". Here, I'll use it a few more times for you: Quantrix was very much like Improv in philosophy, and I would very much like to see a spreadsheet like either one of 'em. The fact that Sun has the OpenStep code-base for Quantrix in their archives gives me some hope that it could happen. Then again, I've always been an optimist.
Sun already owns the rights to Lighthouse Design's application suite. Since these were originally developed for NeXTstep/OpenStep, they should be relatively easy to migrate to Cocoa. I'd sure like to see an Improv/Quantrix like spreadsheet tool put a stake through the heart of Excel!
You know how Europeans always write their dates the reverse of the USA? This is like that, only different.
I disagree that Stallman has attempted to redefine the English language. He himself noted the ambiguity of the word "free", and clarified his intent with the (in)famous quote "Free as in freedom, not as in beer."
Fact 1: MIT has granted Full Professorships to people without degrees. They care about performance and ability more than about degrees.
Fact 2: They also care about integrity. A place like MIT earns and maintains its reputation based on both the quality and the integrity of the work done there. Integrity is where the dean screwed up, and why she is being canned.
I'm fascinated that there's nothing about this on NY Times, CNN, or BBC.
As an MIT alum, I'm a big fan of SICP. However, not many curricula use it and a lot of people "don't get it." I'm not sure why you addressed your response to my earlier message, I wasn't one of the advocates of "learn assembly and C first". I'm saying that Ruby wouldn't be a bad choice for a first language because it doesn't bog you down in lots of overhead early on, but has enough richness that you can carry on with it through some very sophisticated concepts.
You can criticize Ruby for its speed in the various shootouts, but even there I hearken back to the old saying "There are no slow languages, only slow implementations." Any runtime slowness is usually offset by reductions in my development time - I find it easier to prototype an idea in Ruby, then translate to Java/C++ if I need more speed, and for my work Ruby will cough up the answer within a second or two of run time 95% of the time. However, I do academic research with a lot of one-off code - your mileage may vary.
That's still a huge number of people who have exposed themselves to risk.
Not just any compact car. It's a Yugo.
You misspelled "wanker".