4 ms·

I note you failed to actually illustrate species with stable numbers. What the logistic function has to do with anything, I don't know. When reading through the

by waps 13y ago

I note you failed to actually illustrate species with stable numbers. What the logistic function has to do with anything, I don't know. When reading through the reason they choose this function to model population growth it is exactly to construct a stable population growth, not because this was observed in practice. And most certainly the human population curve looks nothing like the logistic function. It was going up much faster than the logistic function predicts. And the current slowdown of human population growth is so much faster than the ramp-up it's ridiculous to compare them. In the logistic function the slowdown mirrors the ramp-up. That's obviously not happening to humans. I can give plenty of popular examples of wildly fluctuating population numbers. Cats versus kangaroos/wallabies versus rabbits versus birds in Australia. Buffalo's versus beavers versus cats versus ... in Yellowstone park. Those are well studied examples of major flips in population density. I also grew up in an area with lots of farms, and I know that for example the number of wolves living in nature follows this pattern. Some years it is so bad they actually attempt to eat adult humans (and sometimes succeed), sometimes there are none. We had a rabbit problem while I was at university, and the biology department investigated. There was an exchange between the number of cats (ex-housecats) and the number of rabbits. First the cats would eat nearly all the rabbits, then they'd starve to the point that few cats remained (and human interference like milk bowls does indeed make things worse), after that the rabbit population multiplied in two years by a factor of more than 1000, after which the cat numbers started growing again. In all these cases you will observe massive swings in the population numbers. The swings were so large that just the swings themselves were at risk of making the species die off entirely (if a swing hits the number 1 or 0, the species never recovers, and has to be reintroduced. In practice that number is 9-10 according to the biology department. For rabbits or cats it never takes long for the species to be reintroduced of course). The biology department paper also claimed that while the timeframes vary wildly from species to species, these swings are the status-quo (within one habitat, the global population is more stable), not the exception. I do know this more from an algorithmic point of view, and I learn biology only as I need it, or if I'm bored or stuck. Which is why I asked you for a few examples of these stable population numbers, as I wouldn't really know. In genetic algorithms stable population numbers don't occur. Normally, because from time t -> t+1 you either let every current try duplicate n -> kn (higher k is better, but requires more resources), which is of course exponential growth. Then you kill all the ones that underperform (which is always basically all of them), or you select the k best performers. Or you can try to be smarter and model sexual reproduction, at which point you grow like n -> k n * (n - 1), which is even faster. There are plenty of variations, but almost all the ones I've ever seen work have exponential growth + 90%+ die offs. Even then genetic algorithms are disappointingly slow in most cases. On the plus side, genetic algorithms will find answers no other algorithm is able to find (much better coverage of the search space). However this comes at a truly massive cost in resources (as you would expect, but it'd be nice if it didn't). So you have an exponential growth, followed by a 99% dieoff (the number that die off has to increase the greater the population size becomes due to memory/cpu constraints).

lutusp 13y ago
> I note you failed to actually illustrate species with stable numbers. In that case, you have no idea what stable numbers means -- it means a species whose long-term average remains the same. Or do you think species are like dormant bank accounts, with constant balances to the penny? > What the logistic function has to do with anything, I don't know. If you don't understand the evolutionary role of the Logistic Function, and since you didn't bother to read the article on that topic I linked, there's no point in this conversation. > And most certainly the human population curve looks nothing like the logistic function. Absolutely false. Do you really want to advertise your ignorance this way? Here are some articles that demonstrate the relation between the human population and the logistic function: http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pdf http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pd... http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://monkeysuncle.stanford.edu/?p=933 http://monkeysuncle.stanford.edu/?p=933 http://www.nature.com/scitable/knowledge/library/how-populations-grow-the-exponential-and-logistic-13240157 http://www.nature.com/scitable/knowledge/library/how-populat... http://arts.bev.net/roperldavid/worldpop.htm http://arts.bev.net/roperldavid/worldpop.htm http://polaris.gseis.ucla.edu/rhayes/courses/Other/Logistic%20Equation.ppt http://polaris.gseis.ucla.edu/rhayes/courses/Other/Logistic%... http://www.jstor.org/discover/10.2307/2981868?uid=3739728&uid=2&uid=4&uid=3739256&sid=21103475204373 http://www.jstor.org/discover/10.2307/2981868?uid=3739728&ui... Just a short sample of the technical articles that, if you would only read them, might alleviate your ignorance. Have a nice day.
- waps 13y ago
  > In that case, you have no idea what stable numbers means -- it means a species whose long-term average remains the same. Or do you think species are like dormant bank accounts, with constant balances to the penny? Given the examples I've seen so far, I'd consider any species that stays within 50% of it's previous numbers over 2 generations stable. But intuitively I'd have a stricter standard, something like no more than 10% per generation. > Absolutely false. Do you really want to advertise your ignorance this way? Heh. Let me just go over your papers : http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pdf http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pd... Essentially makes the case for malthusianism, and proposes to re-introduce population control measures (laws, forced abortions, ...). It's at least 50% a political paper. Need I go on ? I'd really rather not read it in detail. It's also hosted on a site by the social sciences department of the university of Hamburg. Even so, the paper merely intends to "fuel the debate" on this issue, not actually carry out analysis. Allow me to copy a sentence from the paper's conclusion : "Although they [regulatory suggestions to overcome the "normal population cycle"] might seem strangely scandalous and even marginal today, the suggestions and effected regulatory measures to control population growth in the late 1960s and early 1970s merely followed a rigid and conservative accountancy logic still in effect" (I do know a bit of German, the last part is a way-to-literal translation from German. It means to say that while population control laws were repealed, the reasons for having them have become stronger. Now to put things bluntly, but when it comes to scientific credibility, the social sciences rank significantly below my local bookstore's "Fiction" section. I would note that the social sciences see it as their mission not to advance scientific knowledge, but rather to influence public debate and regulation. This obviously means that any social sciences related paper explicitly is NOT the facts, but an attempt to convince you, and should be read like a commercial advertisement. http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://www.cs.xu.edu/math/math120/01f/logistic.pdf A set of exercises constructed to match the logistic function by the math department ... The one example of human population it has actually contradicts your thesis : human population growth starting in 1927, and ending in 2071 (because at that point there would be no more growth) somewhat matches the logistic function. Needless to say this is a prediction. Further analysis reveals that human population growth was calculated at the UN using the logistic function, with the same parameters. As to what this means, I'd have to say it's a very bad example of statistical fitting, nothing more. For example, they just assume they have the answer, and don't even bother to measure the error in their fitting. Please note that this is not criticism of the authors of the paper, who undoubtedly are achieving exactly what they set out to achieve : train students to correctly evaluate a non-trivial calculus function. This is criticism of your interpretation of this paper. http://monkeysuncle.stanford.edu/?p=933 http://monkeysuncle.stanford.edu/?p=933 Is a paper that I would say defends my side of the argument (I'd say it's a critique against the logistic model). Note the prominent line : "Why does the logistic model fail so spectacularly in this case (and many others)?", and look at the graph right above it, and the explanation below. A quote : "Specifically, the logistic model fails to consider mechanisms of population regulation. When density increases, what is affected? Birth rates? Death rates? The r parameter in the logistic model is simply the difference in the gross birth and death rates when there are no conspecifics present." I think it's probably best if I stop now.

by waps 13y ago

lutusp 13y ago

> I note you failed to actually illustrate species with stable numbers. In that case, you have no idea what stable numbers means -- it means a species whose long-term average remains the same. Or do you think species are like dormant bank accounts, with constant balances to the penny? > What the logistic function has to do with anything, I don't know. If you don't understand the evolutionary role of the Logistic Function, and since you didn't bother to read the article on that topic I linked, there's no point in this conversation. > And most certainly the human population curve looks nothing like the logistic function. Absolutely false. Do you really want to advertise your ignorance this way? Here are some articles that demonstrate the relation between the human population and the logistic function: http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pdf http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pd... http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://monkeysuncle.stanford.edu/?p=933 http://monkeysuncle.stanford.edu/?p=933 http://www.nature.com/scitable/knowledge/library/how-populations-grow-the-exponential-and-logistic-13240157 http://www.nature.com/scitable/knowledge/library/how-populat... http://arts.bev.net/roperldavid/worldpop.htm http://arts.bev.net/roperldavid/worldpop.htm http://polaris.gseis.ucla.edu/rhayes/courses/Other/Logistic%20Equation.ppt http://polaris.gseis.ucla.edu/rhayes/courses/Other/Logistic%... http://www.jstor.org/discover/10.2307/2981868?uid=3739728&uid=2&uid=4&uid=3739256&sid=21103475204373 http://www.jstor.org/discover/10.2307/2981868?uid=3739728&ui... Just a short sample of the technical articles that, if you would only read them, might alleviate your ignorance. Have a nice day.

waps 13y ago
> In that case, you have no idea what stable numbers means -- it means a species whose long-term average remains the same. Or do you think species are like dormant bank accounts, with constant balances to the penny? Given the examples I've seen so far, I'd consider any species that stays within 50% of it's previous numbers over 2 generations stable. But intuitively I'd have a stricter standard, something like no more than 10% per generation. > Absolutely false. Do you really want to advertise your ignorance this way? Heh. Let me just go over your papers : http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pdf http://www.neds-projekt.de/Sabine_Hoehler_-_Growth_Curve1.pd... Essentially makes the case for malthusianism, and proposes to re-introduce population control measures (laws, forced abortions, ...). It's at least 50% a political paper. Need I go on ? I'd really rather not read it in detail. It's also hosted on a site by the social sciences department of the university of Hamburg. Even so, the paper merely intends to "fuel the debate" on this issue, not actually carry out analysis. Allow me to copy a sentence from the paper's conclusion : "Although they [regulatory suggestions to overcome the "normal population cycle"] might seem strangely scandalous and even marginal today, the suggestions and effected regulatory measures to control population growth in the late 1960s and early 1970s merely followed a rigid and conservative accountancy logic still in effect" (I do know a bit of German, the last part is a way-to-literal translation from German. It means to say that while population control laws were repealed, the reasons for having them have become stronger. Now to put things bluntly, but when it comes to scientific credibility, the social sciences rank significantly below my local bookstore's "Fiction" section. I would note that the social sciences see it as their mission not to advance scientific knowledge, but rather to influence public debate and regulation. This obviously means that any social sciences related paper explicitly is NOT the facts, but an attempt to convince you, and should be read like a commercial advertisement. http://www.cs.xu.edu/math/math120/01f/logistic.pdf http://www.cs.xu.edu/math/math120/01f/logistic.pdf A set of exercises constructed to match the logistic function by the math department ... The one example of human population it has actually contradicts your thesis : human population growth starting in 1927, and ending in 2071 (because at that point there would be no more growth) somewhat matches the logistic function. Needless to say this is a prediction. Further analysis reveals that human population growth was calculated at the UN using the logistic function, with the same parameters. As to what this means, I'd have to say it's a very bad example of statistical fitting, nothing more. For example, they just assume they have the answer, and don't even bother to measure the error in their fitting. Please note that this is not criticism of the authors of the paper, who undoubtedly are achieving exactly what they set out to achieve : train students to correctly evaluate a non-trivial calculus function. This is criticism of your interpretation of this paper. http://monkeysuncle.stanford.edu/?p=933 http://monkeysuncle.stanford.edu/?p=933 Is a paper that I would say defends my side of the argument (I'd say it's a critique against the logistic model). Note the prominent line : "Why does the logistic model fail so spectacularly in this case (and many others)?", and look at the graph right above it, and the explanation below. A quote : "Specifically, the logistic model fails to consider mechanisms of population regulation. When density increases, what is affected? Birth rates? Death rates? The r parameter in the logistic model is simply the difference in the gross birth and death rates when there are no conspecifics present." I think it's probably best if I stop now.