|
Post by Eva Yojimbo on Feb 19, 2020 3:19:36 GMT
I have no socks. Admin has my permission to look at my IP and tell you if I have any socks. You're being accused of having socks? Good lord, no one is safe... I knew it before I looked: Your IP doesn't match any other account here, so if you have any socks, I'm not aware of them. Case dismissed. Thanks. Arlon suspected gadreel was my sock.
|
|
|
Post by Admin on Feb 19, 2020 3:33:03 GMT
You're being accused of having socks? Good lord, no one is safe... I knew it before I looked: Your IP doesn't match any other account here, so if you have any socks, I'm not aware of them. Case dismissed. Thanks. Arlon suspected gadreel was my sock. Ditto for gadreel. This proving innocence business is becoming so very commonplace, we might have to actually change the format. What a wonderful world that would be.
|
|
|
Post by Eva Yojimbo on Feb 19, 2020 4:10:31 GMT
Arlon10Since now you're just going to claim that I've copied the Bayesian solution to the Monty Hall Problem from others, I created a different problem to solve here: imdb2.freeforums.net/thread/211197/hairy-ballsy-problem Feel free to solve it yourself. I already have using Bayes (I sent the solution in a PM to AJ_June so my solution is timestamped). You can't accuse me of ripping anyone off to solve that one since, AFAIK, that problem is an invention of mine.
|
|
|
Post by Admin on Feb 19, 2020 7:31:03 GMT
Arlon10Since now you're just going to claim that I've copied the Bayesian solution to the Monty Hall Problem from others, I created a different problem to solve here: IMDB2.freeforums.net/thread/211197/hairy-ballsy-problem Feel free to solve it yourself. I already have using Bayes (I sent the solution in a PM to AJ_June so my solution is timestamped). You can't accuse me of ripping anyone off to solve that one since, AFAIK, that problem is an invention of mine. About the Monty Hall problem, I don't know nothing about no Bayer dudes, but... 1: You pick the car, the host shows a goat, you switch, you lose. 2: You pick Goat 1, the host shows Goat 2, you switch, you win. 3: You pick Goat 2, the host shows Goat 1, you switch, you win. In 2/3 of these scenarios, you win if you switch and lose if you don't. Seems like a no-brainer to me. On the other hand, if there's another possible scenario I'm not seeing, then I'm the no-brainer in this situation and I wasn't here and you never saw me.
|
|
|
Post by Arlon10 on Feb 19, 2020 9:14:39 GMT
How many years have you here counting the previous boards? Verbs facilitate meaning in sentences. How many years have you been here counting the previous boards?
|
|
|
Post by Arlon10 on Feb 19, 2020 9:23:32 GMT
I have no socks. Admin has my permission to look at my IP and tell you if I have any socks. You're being accused of having socks? Good lord, no one is safe... I knew it before I looked: Your IP doesn't match any other account here, so if you have any socks, I'm not aware of them. Case dismissed. He can't have two IPs? Many people have more.
|
|
|
Post by Arlon10 on Feb 19, 2020 9:33:33 GMT
Arlon10 Since now you're just going to claim that I've copied the Bayesian solution to the Monty Hall Problem from others, I created a different problem to solve here: IMDB2.freeforums.net/thread/211197/hairy-ballsy-problem Feel free to solve it yourself. I already have using Bayes (I sent the solution in a PM to AJ_June so my solution is timestamped). You can't accuse me of ripping anyone off to solve that one since, AFAIK, that problem is an invention of mine. I'm sure it is since the English is so bad I can't tell what you want to know and don't want to find out.
|
|
|
Post by Admin on Feb 19, 2020 9:38:08 GMT
You're being accused of having socks? Good lord, no one is safe... I knew it before I looked: Your IP doesn't match any other account here, so if you have any socks, I'm not aware of them. Case dismissed. He can't have two IPs? Many people have more. In my courtroom, proof of guilt is required for a conviction.
|
|
|
Post by Arlon10 on Feb 19, 2020 9:46:48 GMT
This is not my job. It is not my hobby. I do not believe it has real value. Obviously you spend far more time on it than I do. I suspect because you think it has more value than it does, or you gamble too much. The defenders of Bayes' Theorem that agreed with your overrating of it are gone. That is except "gadreel" whom I have long suspected is your sock. AJ has to study it for his goals in school, but I think he realizes that real life does not employ the theorem very often. The taxicab problem is not real and it shows terribly. If taxicab companies are really different and they use the same makes and models of vehicles and you can't tell one from another they would add markings or equipment of some kind so that you can tell. That's just common sense. A statistical assessment adequate to be meaningful of only one person's color detection is going to be beyond the time, expense and usefulness for any court's resources. It would probably show a binary condition anyway, can or can't distinguish the colors. The taxicab problem is for people with no common sense. Again you do not show your work when you list the theorem's "historic" accomplishments. I gamble for a living, so it's quite valuable to me, and, again, it was quite valuable in two wars. I have no socks. Admin has my permission to look at my IP and tell you if I have any socks. The taxi cab problem shows how people under/overestimate priors and/or new evidence. Its purpose is to show the existence of cognitive biases that deviate from rationality. Kahneman has dozens of examples of this in his books on cognitive bias. How would I "show my work" when listing the theorem's historic accomplishments? Do you know nothing about how the Enigma code was cracked? You can read about them HERE, but there's nothing I can do to overcome your desire not to believe facts and science you dislike. Here are several links that discuss Bayes and The Enigma Code (the first discusses many more applications): theconversation.com/bayes-theorem-the-maths-tool-we-probably-use-every-day-but-what-is-it-76140 www.intelligentinvestor.com.au/investment-news/the-theory-that-cracked-the-enigma-code/138342www.singingbanana.com/enigmaproject/maths.pdfYou can treat any series of events as a single event probabilistically. Just multiply the probabilities of each event happening together. However, this is not necessary in the 100-door problem, because asking the probability of any door being left shut (1/99) gives us the same probability of opening any sequence of 98 doors. If you were asking about any particular sequence's probability (say, he opens 77, then 43, then 2, then 87, etc.) that would be a different matter, but we don't care about the particular sequence of opening. It seems you slipped from talking about the 99-door variation to the 3-door variation. I also don't know why you're making "not opening door 3" "Event B." First off, the "event" is the opening of door 2, not the "not opening" of door 3. I mean, I guess they're technically equivalent, but I have no idea why you'd prefer to write that as a double negative... and you complain about my English? Second, there is no "Event A." "A" is a prior, not an event. Finally, the only reason your solution works in this case is because the priors happen to be 50/50. If the priors were anything else, your "P=1 behind D3, subtract 1/3" wouldn't work." The reason it's actually .66 is because the priors for both D3 and D1 are 50/50. When you multiple D3 by 1 (as you did), and multiple D1 by .5 (which you didn't), you end up with 50/25, and because 50 is twice 25, it's .67/.33. Amazing how everyone solving Monty Hall using Bayes is not giving a "proper solution" and only you are. I guess when the Professor at the Department of Statistics at the University of Toronto is also solving Monty Hall using Bayes, he still doesn't know what he's talking about. What would a professor of Statistics at a major University know about probability? I think I'm going to concoct my own little problem that you can't solve using your "proper solution." We'll see how well you do. I'm sorry I gave away an actual solution to the Monty Hall problem. I doubt many people will find it before it scrolls away though. Meanwhile I must say you are doing a great job of discouraging people from finding answers on the internet, copying and pasting for their homework. You truly make the internet a PoS. My concern is that people who believe what you and your associates say are voting and that is ruining the country. An uninformed or misinformed populace will ruin democracy. You are not my teacher obviously. If my guess is correct you don't really want to be. If you don't really want to be, rest assured that will never happen.
|
|
|
Post by Admin on Feb 19, 2020 9:48:48 GMT
I'm sorry I gave away an actual solution to the Monty Hall problem. I doubt many people will find it before it scrolls away though. I'd like to. Got a link? Or are you going to make me read this thread?
|
|
|
Post by Arlon10 on Feb 19, 2020 9:49:19 GMT
He can't have two IPs? Many people have more. In my courtroom, proof of guilt is required for a conviction. There is always suspicion, more or less.
|
|
|
Post by Arlon10 on Feb 19, 2020 9:51:47 GMT
I'm sorry I gave away an actual solution to the Monty Hall problem. I doubt many people will find it before it scrolls away though. I'd like to. Got a link? Or are you going to make me read this thread? One of us will have to read the thread either again or for the first time. It's in there somewhere.
|
|
|
Post by Admin on Feb 19, 2020 10:03:28 GMT
In my courtroom, proof of guilt is required for a conviction. There is always suspicion, more or less. You would convict on suspicion? That's scary. I'd like to. Got a link? Or are you going to make me read this thread? One of us will have to read the thread either again or for the first time. It's in there somewhere. In that case, it would seem that it did indeed get lost. That's a pity, but I assume it's probably just a longer, more complicated version of mine.
|
|
|
Post by Eva Yojimbo on Feb 19, 2020 10:48:15 GMT
I gamble for a living, so it's quite valuable to me, and, again, it was quite valuable in two wars. I have no socks. Admin has my permission to look at my IP and tell you if I have any socks. The taxi cab problem shows how people under/overestimate priors and/or new evidence. Its purpose is to show the existence of cognitive biases that deviate from rationality. Kahneman has dozens of examples of this in his books on cognitive bias. How would I "show my work" when listing the theorem's historic accomplishments? Do you know nothing about how the Enigma code was cracked? You can read about them HERE, but there's nothing I can do to overcome your desire not to believe facts and science you dislike. Here are several links that discuss Bayes and The Enigma Code (the first discusses many more applications): theconversation.com/bayes-theorem-the-maths-tool-we-probably-use-every-day-but-what-is-it-76140 www.intelligentinvestor.com.au/investment-news/the-theory-that-cracked-the-enigma-code/138342www.singingbanana.com/enigmaproject/maths.pdfYou can treat any series of events as a single event probabilistically. Just multiply the probabilities of each event happening together. However, this is not necessary in the 100-door problem, because asking the probability of any door being left shut (1/99) gives us the same probability of opening any sequence of 98 doors. If you were asking about any particular sequence's probability (say, he opens 77, then 43, then 2, then 87, etc.) that would be a different matter, but we don't care about the particular sequence of opening. It seems you slipped from talking about the 99-door variation to the 3-door variation. I also don't know why you're making "not opening door 3" "Event B." First off, the "event" is the opening of door 2, not the "not opening" of door 3. I mean, I guess they're technically equivalent, but I have no idea why you'd prefer to write that as a double negative... and you complain about my English? Second, there is no "Event A." "A" is a prior, not an event. Finally, the only reason your solution works in this case is because the priors happen to be 50/50. If the priors were anything else, your "P=1 behind D3, subtract 1/3" wouldn't work." The reason it's actually .66 is because the priors for both D3 and D1 are 50/50. When you multiple D3 by 1 (as you did), and multiple D1 by .5 (which you didn't), you end up with 50/25, and because 50 is twice 25, it's .67/.33. Amazing how everyone solving Monty Hall using Bayes is not giving a "proper solution" and only you are. I guess when the Professor at the Department of Statistics at the University of Toronto is also solving Monty Hall using Bayes, he still doesn't know what he's talking about. What would a professor of Statistics at a major University know about probability? I think I'm going to concoct my own little problem that you can't solve using your "proper solution." We'll see how well you do. I'm sorry I gave away an actual solution to the Monty Hall problem. I doubt many people will find it before it scrolls away though. Meanwhile I must say you are doing a great job of discouraging people from finding answers on the internet, copying and pasting for their homework. You truly make the internet a PoS. My concern is that people who believe what you and your associates say are voting and that is ruining the country. An uninformed or misinformed populace will ruin democracy. You are not my teacher obviously. If my guess is correct you don't really want to be. If you don't really want to be, rest assured that will never happen. LMAO that you "gave it away." The solution has been out there for years, long before you and I started discussing it. For your sake (and for Admin's) I will post both of our solutions so they will be easy to find. You can lie and I say I "copy and pasted" all you want. I didn't. I've known about Bayes for a long time, long before I read/heard about Monty Hall, and when I did discover Monty Hall it was a straight-forward application for me. You can also "suspect" me of having socks. I do not. This is my one and only account here and on every forum I've ever been on. I take no delight in trolling or in playing "characters" other than myself on message boards. If I wanted to play someone other than myself, I'd start up a local D&D game or take an improv/acting class.
|
|
|
Post by Eva Yojimbo on Feb 19, 2020 10:49:59 GMT
Arlon10 Since now you're just going to claim that I've copied the Bayesian solution to the Monty Hall Problem from others, I created a different problem to solve here: IMDB2.freeforums.net/thread/211197/hairy-ballsy-problem Feel free to solve it yourself. I already have using Bayes (I sent the solution in a PM to AJ_June so my solution is timestamped). You can't accuse me of ripping anyone off to solve that one since, AFAIK, that problem is an invention of mine. I'm sure it is since the English is so bad I can't tell what you want to know and don't want to find out. Where/How is the "English so bad" you can't solve it? The questions are straight forward: Is one option more probable than the others? If so, how much more probable? If there's anything unclear, you can ask me to clarify. I suspect you will not because you can't solve it. Hint: Bayes can.
|
|
|
Post by Eva Yojimbo on Feb 19, 2020 11:16:23 GMT
Arlon10 Since now you're just going to claim that I've copied the Bayesian solution to the Monty Hall Problem from others, I created a different problem to solve here: IMDB2.freeforums.net/thread/211197/hairy-ballsy-problem Feel free to solve it yourself. I already have using Bayes (I sent the solution in a PM to AJ_June so my solution is timestamped). You can't accuse me of ripping anyone off to solve that one since, AFAIK, that problem is an invention of mine. About the Monty Hall problem, I don't know nothing about no Bayer dudes, but... 1: You pick the car, the host shows a goat, you switch, you lose. 2: You pick Goat 1, the host shows Goat 2, you switch, you win. 3: You pick Goat 2, the host shows Goat 1, you switch, you win. In 2/3 of these scenarios, you win if you switch and lose if you don't. Seems like a no-brainer to me. On the other hand, if there's another possible scenario I'm not seeing, then I'm the no-brainer in this situation and I wasn't here and you never saw me. Bayes was a statistician who invented a theorem that can solve problems like this. Your solution is basically using Bayes on an intuitive level. Another way to word your solution in a more precisely Bayesian way is: 1. If you picked the car (33% of the time), the host shows one of two (50%) goats = .33*.5 = .165 2. If you picked the goat not shown (33% of the time), the host always (100%) shows the other goat = .33*1 = .33 If you divide the second possibility (.33) by the total of both possibilities (.5), it's .67. Meaning that 2/3 of the time you've picked a goat and should switch. Shorter methods like the one you used can work for a simple problem like this, not so well when problems get more complex, but Bayes can apply to any situation where conditional reasoning/probability is used to update prior probabilities. I'm sorry I gave away an actual solution to the Monty Hall problem. I doubt many people will find it before it scrolls away though. I'd like to. Got a link? Or are you going to make me read this thread? Here's a brief timelines of mine and Arlon's discussion: Opening salvo (I'm referencing an old discussion we had on this in a different thread): imdb2.freeforums.net/post/3663605/threadArlon's first attempt at solving/explaining it: imdb2.freeforums.net/post/3667349/threadMy solution using long-form Bayes: imdb2.freeforums.net/post/3672877/threadAlso, since we're (Arlon and I) are just two random dudes on an internet forum, I've linked (several times) to Jeffrey S. Rosenthal's (professor at the Department of Statistics at Toronto University) showing the importance of Bayes to the Monty Hall problem and variations on the problem: www.probability.ca/jeff/writing/montyfall.pdf
|
|
|
Post by general313 on Feb 19, 2020 16:10:26 GMT
I do think science depends on a small kernel of philosophy (if done right), to provide a frameworks sort of like the axioms in math and geometry. I think it should be minimized, just as reducing axioms to a bare minimum is an aim in math. I think of it as a bootstrap program used to boot up the full operating system for a computer. The bootstrap is made as small an simple as possible, as, especially in the early days of computing, it had to be entered manually via the computer's front panel. I may have to check out Jayne's book. Do he or she have a good writing style? I think Einstein's resistance to QM and Copenhagen in particular was much more than an insecurity about his theories of relativity. It seems to have been a deep-seated aesthetic disgust at the concept of indeterminacy. Even in his later years, after there were multiple and very solid confirmations of relativity, he remained opposed to Copenhagen, and in fact spent much of his later life trying to unify QM and relativity (unsuccessfully). It would be nice to know Einstein's reaction to MW. To me what is really weird (and really stretches credulity) about MW is that it seems to be a violation of conservation of energy. In ordinary everyday experience, any time we duplicate something, whether in the physical world or in software, extra resources are required to support the increased count of instances (material if physical, computer memory if in software). In a MW scenario (even if all the duplications happen at the beginning), to support the apparent astronomical number of parallel worlds would imply the existence of an astronomical amount of energy. This may be possible but it's quite mind-blowing. Yet if the extent of the observable universe as we understand it today were presented to the ancient Greeks, their minds would have been blown too. So who knows? By the way, interesting discussion! I essentially agree about reducing/minimizing philosophy in science. That's what I like about Bayes; it's simple. I honestly don't believe anything else is needed, philosophy-wise. Hypotheses are priors, experiments/evidence are conditionals, theories are what happens when the experiments/evidence make a hypothesis overwhelmingly likely. For the Jaynes book, I must stress it's meant as a textbook on probability theory, the kind of thing you'd read in college if you were taking a course of probability. When Jaynes is using examples it's perfectly readable, but there are also parts that are quite technical. I found it easy enough to skip over the confusing/technical bits, but still follow along with most of the examples. I don't think it's a book where one must understand every single page to get the substance; understanding every page would only be necessary if one was actually planning on becoming fully educated in probability theory. I think it's mostly interesting to see how he applies the same fundamental reasoning from everything to the most technical examples in science to everyday reasoning about ordinary things. The conservation of energy is a common complaint about MW, but it's unfounded. The reason being that nothing's actually being duplicated. When particles are in a state of superpositioning, that means they're already in multiple places at once. We see this in the double-slit experiment where even firing single particles at a double-slit shows a wave-pattern of single particles interfering with themselves. All of those states--say, a particle being "spin-up" and "spin-down"--are "worlds." When we measure, we (or our measuring devices) are also in multiple states. Our multiple states entangle with the multiple states of the particles, and they decohere from each other into the rest of the environment, essentially separating. Nothing is "duplicated" or "created" in this process; it's just multiple states of multiple systems entangling and decohering. A good overview is from the physicist Sean Carroll: www.preposterousuniverse.com/blog/2014/06/30/why-the-many-worlds-formulation-of-quantum-mechanics-is-probably-correct/What other interpretations like Copenhagen and Bohm do is that they add stuff to QM to essentially eliminate those other states. Copenhagen adds a stochastic collapse, so that observation causes the other states to go "poof" and disappear. How? Why? Nobody knows. Bohm adds a hidden variable, a kind of "pilot-wave" that "selects" one of the "worlds" so that the other "worlds" are empty. How? Why? Nobody knows. MWI just does away with those additional assumptions. There is, however, a legitimate criticism of MWI, and Sean Carroll addresses this, and it's this: where do the probabilities come from? The Born Rule is where all the predictive power of QM comes from, it tells us probabilities are given by the wavefunction squared. In MWI it's not clear why this is so, while in Shrodinger and Bohm the probabilities are just an innate part of wavefunction. Interesting discussion indeed! As I was thinking about the energy "problem" and the computer simulation equivalent, where multiple copies of an object or thread consume more and more computing resources, it occurred to me that a quantum computer wouldn't have this problem. It would come naturally to a quantum computer to handle a multiplicity of threads simultaneously, with no change in the load/demand on computing resources. Maybe the universe is a large quantum computer? Maybe it will be possible to start out with an assumption-less version of QM, apply some knowledge learned from quantum computing to arrive at a MW scenario.
|
|
|
Post by general313 on Feb 19, 2020 16:22:46 GMT
In my courtroom, proof of guilt is required for a conviction. There is always suspicion, more or less. Why on earth you would imagine Eva and Gadreel of being the same person is beyond me. Their styles, talents and interests are nothing like each other, and even their posting schedules show a completely different pattern. And, if I'm not mistaken, one is American and the other from somewhere down under. Now maybe one person has devised an elaborate hoax, carefully and meticulously crafting the illusion of distinct individuals posting here, for the purpose of undermining your arguments about Bayes' theorem!
|
|
|
Post by gadreel on Feb 19, 2020 18:54:42 GMT
In my courtroom, proof of guilt is required for a conviction. There is always suspicion, more or less. You mean like I suspect that you should not be allowed out of your house without adult supervision?
|
|
|
Post by goz on Feb 19, 2020 20:45:32 GMT
He can't have two IPs? Many people have more. In my courtroom, proof of guilt is required for a conviction. Well said admin, and with this poster 'proof' of anything seems to his Achilles 'heel'!
|
|