Re: [FRIAM] Real-world genetic algorithm example... help!
Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the case of individual chickens, the fitness function was ill-defined and didn't work very well. In particular, this section points out that it is not necessary to know *why *a good solution is good. *Why *doesn't have to come into it ... the fitness function simply ensures that the best solution, no matter what the reasons are for being the best, can emerge from this process. In regards to Russ' complaints, I'm not sure I can agree that no crossover/mutation occurred. I haven't read the original paper yet, just the Huff Post treatment, so I didn't realize that the chicken clusters weren't mixed. That is, I just assumed that more than one cluster was selected among the best, and that they collectively produced the subsequent generations. However, consider the case of mutation. Russ says there is no mutation within the population elements - the clusters of chickens. But functionally, there actually is mutation. This becomes obvious when we remember that a second-generation chicken coop is different from the first-generation coop. The genes were all there, but some of them weren't expressed ... that is, they simply combined together in a different way to produce a different coop. It doesn't matter that the kids have all the genes of the parents ... the kids are still different. And we know this is true because egg production went up. This couldn't have happened unless there was *something *(crossover or mutation) that changed from generation to generation. Regarding James' point, I don't know how the roosters were handled from generation to generation (something that is probably in the original paper). But I suppose they could get the next generation roosters the same way they got the next generation hens - by simply hatching a few eggs. One final point: since GA originally got its inspiration from biology, I see no reason why biology can be used to illustrate GA in a textbook. Thoughts? Cheers, -Ted On Fri, Jul 9, 2010 at 10:03 PM, ERIC P. CHARLES e...@psu.edu wrote: Russ, Completely agreed. I'm not sure how one would connect the chicken stuff in a pretty way to standard computer genetic algorithms. I suppose one could relate them together to suggest the need for variation in selection methods when using GAs. That's Ted's part. I only claimed to know how the chicken part worked through (either artificial or natural) selection for something other than best individual production. Eric On Fri, Jul 9, 2010 09:18 PM, *Russ Abbott russ.abb...@gmail.com* wrote: It's a great story, but it's not a genetic algorithm as we normally think about it. It's really just breeding. For one thing, no computer was involved. The point of the whole thing is to establish the notion of group selection, which was forbidden in the biological world for a while. This experiment shows that it makes sense. In what sense was it just breeding? Well, what was bred was coops rather than chickens. So the original population was 6 coops. The best one was selected and propagated. The best of those was selected, etc. Not at all what GA is about. There was no crossover or mutation between the population elements -- which are coops. Of course there is crossover among the chickens in the coop, but it wasn't chickens that were bred. The fitness function was a function applied to the coop. So even though it is a very nice experiment and even though it makes a very strong case for group selection, it's probably not a good example for a chapter on genetic algorithms in a text book. -- Russ On Fri, Jul 9, 2010 at 4:25 PM, ERIC P. CHARLES e...@psu.edu#129ba189d8601434_ wrote: Shawn, The two ways to answer your question would either be to invoke artificial selection (i.e., because you can design a genetic algorithm to do anything you want, just as chicken breeders can keep whichever eggs or to invoke Wilson's trait group selection. In trait group selection you break selection into two parts, within-group and between-group selection. If you do that, you can, under the right conditions, find that types of individuals who reproduce less well within any group can still out-compete the competition when you look between groups. Math available upon request. I have a vague memory that this has come across the FRIAM list before. Eric On Fri, Jul 9, 2010 06:47 PM, *Shawn Barr sba...@gmail.com#129ba189d8601434_ * wrote: Ted, I'm confused. Why would a genetic algorithm ever select a hen that produces fewer eggs over a hen that produces more eggs? Shawn On Fri, Jul 9, 2010 at 2:57 PM, Ted Carmichael teds...@gmail.com#129ba189d8601434_129b987e5d851537_ wrote: Nick, this is perfect. Thank you! BTW - the reason for this request is, my advisor and I
Re: [FRIAM] Need system-oriented Java-OO developers, modeling/sim environment, part-full time, work-at-home
Wow ... it is a small world sometimes. We started talking about Genetic Algorithms on a different thread. So of course I was thinking about Dr. Michalewicz, who taught a class in GA that I took a few years ago, and (literally) wrote the book on the subject. And then the very next thread I read is from Grant, soliciting work for the company that Dr. M. founded. How very auspicious. Grant, if you'd like to pass along the note to students/recent graduates from UNC Charlotte, I could put you in touch with the right people, I'm sure. Cheers, -Ted On Thu, Jul 8, 2010 at 1:33 PM, Grant Holland grant.holland...@gmail.comwrote: Friends, I'm looking for a couple of system-oriented Java developers, object-oriented, for a complex project involving modeling and simulation (mixed ABM and discrete-event) environment. Some middleware-level Java optionally dev involved. Software engineering best practices involved. Mostly work-at-home, occasional online meetings, little (if any) travel. Contact me via return email or cell. Thanks, Grant -- Grant Holland VP, Product Development and Software Engineering NuTech Solutions 404.427.4759 FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Real-world genetic algorithm example... help!
It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the case of individual chickens, the fitness function was ill-defined and didn't work very well. In particular, this section points out that it is not necessary to know *why *a good solution is good. *Why *doesn't have to come into it ... the fitness function simply ensures that the best solution, no matter what the reasons are for being the best, can emerge from this process. In regards to Russ' complaints, I'm not sure I can agree that no crossover/mutation occurred. I haven't read the original paper yet, just the Huff Post treatment, so I didn't realize that the chicken clusters weren't mixed. That is, I just assumed that more than one cluster was selected among the best, and that they collectively produced the subsequent generations. However, consider the case of mutation. Russ says there is no mutation within the population elements - the clusters of chickens. But functionally, there actually is mutation. This becomes obvious when we remember that a second-generation chicken coop is different from the first-generation coop. The genes were all there, but some of them weren't expressed ... that is, they simply combined together in a different way to produce a different coop. It doesn't matter that the kids have all the genes of the parents ... the kids are still different. And we know this is true because egg production went up. This couldn't have happened unless there was *something *(crossover or mutation) that changed from generation to generation. Regarding James' point, I don't know how the roosters were handled from generation to generation (something that is probably in the original paper). But I suppose they could get the next generation roosters the same way they got the next generation hens - by simply hatching a few eggs. One final point: since GA originally got its inspiration from biology, I see no reason why biology can be used to illustrate GA in a textbook. Thoughts? Cheers, -Ted On Fri, Jul 9, 2010 at 10:03 PM, ERIC P. CHARLES e...@psu.edu wrote: Russ, Completely agreed. I'm not sure how one would connect the chicken stuff in a pretty way to standard computer genetic algorithms. I suppose one could relate them together to suggest the need for variation in selection methods when using GAs. That's Ted's part. I only claimed to know how the chicken part worked through (either artificial or natural) selection for something other than best individual production. Eric On Fri, Jul 9, 2010 09:18 PM, *Russ Abbott russ.abb...@gmail.com*wrote: It's a great story, but it's not a genetic algorithm as we normally think about it. It's really just breeding. For one thing, no computer was involved.
Re: [FRIAM] Real-world genetic algorithm example... help!
Selecting for productive coops rather than productive hens might reject highly productive, highly aggressive hens in favor of somewhat less productive, considerably less aggressive hens who would leave their coop-mates in peace (and therefore able to produce more eggs). Such hens need not have anything like a “concept” of loyalty to the coop --we could redistribute these hens to different coops and without affecting coop productivity. But after a while, we might find we are selecting for highly productive, potentially highly aggressive hens who are strongly inhibited against bothering a hen they grew up with. Then if we redistributed the hens among different coops, coop productivity would decrease. Are there applications to genetic algorithms? It shows you have to be careful about dividing the task to be done into subtasks. You don’t want to overlook an algorithm for doing one subtask that provides useful byproducts for another subtask. Instead of selecting for each subtask separately, you might select for teams of algorithms that do the whole task. From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 2:50 AM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Real-world genetic algorithm example... help! It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.commailto:teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the case of individual chickens, the fitness function was ill-defined and didn't work very well. In particular, this section points out that it is not necessary to know why a good solution is good. Why doesn't have to come into it ... the fitness function simply ensures that the best solution, no matter what the reasons are for being the best, can emerge from this process. In regards to Russ' complaints, I'm not sure I can agree that no crossover/mutation occurred. I haven't read the original paper yet, just the Huff Post treatment, so I didn't realize that the chicken clusters weren't mixed. That is, I just assumed that more than one cluster was selected among the best, and that they collectively produced the subsequent generations. However, consider the case of mutation. Russ says there is no mutation within the population elements - the clusters of chickens. But functionally, there actually is mutation. This becomes obvious when we remember that a second-generation chicken coop is different from the first-generation coop. The genes were all there, but some of them weren't expressed ... that is, they simply combined together in a different way to produce a different coop. It doesn't matter that the kids have all the genes of the parents
Re: [FRIAM] Real-world genetic algorithm example... help!
Well, in regards to (1), yes, I would guess elitism + mutation is a good description. However, I believe that is enough to qualify as a GA. As I recall, some GA practitioners believe mutation is best, some believe crossover is best, and some feel you should have both, or decide based on the problem. I would guess that it is a minority viewpoint to claim that mutation by itself is not enough to be a GA. For (3), you say that the genetic operators are not explicit, in the chicken coop example. My understanding is that sometimes the genetic operator itself is a part of the genotype, and is thus subject to mutation/crossover as well. In such a case, it wouldn't really be explicit because it would vary across the population. Of course, the final genetic operator is discoverable - i.e., it is recorded in the final solution ... but this doesn't really matter to the programmer. He has no knowledge *a priori* of what that operator will turn out to be; and further, the final genetic operator is not the point of the GA. Finding a good solution is the desired outcome - the operator is secondary at best. For (2), you correctly imply that a phenotype - or genotype - *must *be explicitly defined in a computer. Well, sure. The computer is deterministic, and so this information will be recorded somewhere as part of the code of the solution. What I find interesting is this idea that the programmer has to know and care and understand the final solution. I don't think that is the case. Oh sure, in some instances the final solution is quite clear. For example, the TSP will end up with a list of city-pairs that is easily understood. But I can also imagine instances of a GA, say applied to computer code, or a mathematical formula, that becomes so immensely complex that the researcher does not understand why the final solution works. And, as pointed out above, he doesn't have to understand why it works. That it does work well is good enough, isn't it? I just don't see any functional distinction between not caring why the final solution works and - in the case of chickens - not being able to precisely describe *how *it works or what it looks like. It's kind of like that DARPA funded robot pack-animal http://www.bostondynamics.com/robot_bigdog.html... they don't care, really, what all the final rules are, as long as the robot can walk. In fact, I would suspect that some flexibility in the 'final solution' is allowed, and that the machine learning process is continuously running to some degree. But I am enjoying the discussion, so thanks for that. Cheers, -Ted On Sat, Jul 10, 2010 at 2:50 AM, Russ Abbott russ.abb...@gmail.com wrote: It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the
Re: [FRIAM] Projects: 5 Stages
Tory -- It was mostly that the stages seem to be empirically valid - I can recall many instances where I've been in a team or relationship that had the excitement and novelty of coming together, the inevitable misunderstandings/arguments about how to proceed, a reconciliation and synthesis of the preferred approach, and finally, working together along those preferred lines to achieve something. The hazard is that it's presented in such a way as to suggest that there's an orderly, linear progression, whereas we know it's often quite the contrary. I tend to see it (and, relating it to my thinking) as a continuum - the progression is a road map, perhaps, but I'm likely to be taking on ramps and off ramps along the way. It's not at all clear, either in my own mind or when I'm working with others, where the transitions occur: no bright line between storming and norming, for example. - Claiborne - -Original Message- From: Victoria Hughes victo...@toryhughes.com To: The Friday Morning Applied Complexity Coffee Group friam@redfish.com Sent: Fri, Jul 9, 2010 11:00 pm Subject: Re: [FRIAM] Projects: 5 Stages Dunno. Not familiar with that. One aim of mine with this book is to phrase these ideas in a way that the beloved General Public can use them. Not just B-school types. I want the basic concept to be generally accessible. Needs to be, after all. Will look into this. Has it affected how you conceptualize and take action on ideas and goals? Or was it interesting (partly because of the alliteration, that memorable lilting he set up sticks in our brains like the Oscar Mayer Weiner song) Happy to hear speculations, no worries. Tory On Jul 9, 2010, at 6:40 PM, q...@aol.com wrote: Tory -- How does this relate (if at all) to the simplistic group dynamics model I learned in business school (attributed to Bruce Tuckman)? forming storming norming performing At a minimum, I'm missing a stage, and I'm sure there's much more to your analysis. Excuse my speculations. - Claiborne Booker - -Original Message- From: Victoria Hughes victo...@toryhughes.com To: The Friday Morning Applied Complexity Coffee Group friam@redfish.com Sent: Fri, Jul 9, 2010 8:14 pm Subject: Re: [FRIAM] Projects: 5 Stages Yup, in most cases. Sometimes limitations force unusual, possibly more successful, resolutions. I don't know the book, will look into it. Thanks. Tory On Jul 9, 2010, at 5:51 PM, Stephen Thompson wrote: Tory: I am part way through Scott Page's book titled The Difference He discussed the the power of diversity to produce better groups and outcomes. Are you aware of that reference? None, some, or much diversity would influence the stages or at least successful completion of the stages would it not? Steph T Victoria Hughes wrote: Fascinating. The original story and its appearance/discussion here. I am writing a book on the five simple stages that projects move through, from idea to reality. Part of the chapter, whose midst I am in, discusses teams, inner and outer: the grouping of abilities and attributes required to get unstuck and get something done. Sometimes the 'crate o' chickens' is outside of us, if we are working with a team. Sometimes our team is made from aspects of our own mind: the internal - complex- interconnection of knowledge, abilities, ideas, etc all squawking, laying, attacking, defending, at once, inside our brains. Glad to know that even among the inheritors of the reptilian hind brain there can be cooperation for a larger good, even if that is for more chickens. Tory On Jul 9, 2010, at 4:53 PM, Ted Carmichael wrote: Well, it wouldn't ... unless you were selecting for the lowest producing hens. The GA selects for the groups of chickens that produce the most eggs, not the individuals. Some of those individuals may actually not produce many eggs, but they must somehow help the ones that do produce more eggs (in their group). -t On Fri, Jul 9, 2010 at 6:47 PM, Shawn Barrsba...@gmail.com wrote: Ted, I'm confused. Why would a genetic algorithm ever select a hen that produces fewer eggs over a hen that produces more eggs? Shawn On Fri, Jul 9, 2010 at 2:57 PM, Ted Carmichael teds...@gmail.com wrote: Nick, this is perfect. Thank you! BTW - the reason for this request is, my advisor and I were asked to write a chapter on Complex Adaptive Systems, for a cognitive science textbook. In it, I talk briefly about GA, and put this story about the chickens in because I thought it was a neat example. I'll add the references now. Much appreciated. -t
Re: [FRIAM] Real-world genetic algorithm example... help!
Everybody, Why do the best conversations happen when I am totally unable to pay proper attention to them!? Somebody help me out here. A genetic algorithm is a PROCEDURE, right? So you run the procedure on a computer. Is it possible to implement that same procedure on crates of chickens. In Gallinacea so to speak? Let the group of chickens compute the algorithm? Total misuse of language? Take groups of chickens, raise them in crates of nine. The crate is the individual; the individual chickens are the germ cells. Count the number of eggs produced by the crates. Hatch the eggs produced by the crate with the most eggs. Raise the chicklets. Put them in crates of nine. Etc. Hot as hell here and hard to THINK. n Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] - Original Message - From: Russ Abbott To: The Friday Morning Applied Complexity Coffee Group Sent: 7/10/2010 2:51:13 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the case of individual chickens, the fitness function was ill-defined and didn't work very well. In particular, this section points out that it is not necessary to know why a good solution is good. Why doesn't have to come into it ... the fitness function simply ensures that the best solution, no matter what the reasons are for being the best, can emerge from this process. In regards to Russ' complaints, I'm not sure I can agree that no crossover/mutation occurred. I haven't read the original paper yet, just the Huff Post treatment, so I didn't realize that the chicken clusters weren't mixed. That is, I just assumed that more than one cluster was selected among the best, and that they collectively produced the subsequent generations. However, consider the case of mutation. Russ says there is no mutation within the population elements - the clusters of chickens. But functionally, there actually is mutation. This becomes obvious when we remember that a second-generation chicken coop is different from the first-generation coop. The genes were all there, but some of them weren't expressed ... that is, they simply combined together in a different way to produce a different coop. It doesn't matter that the kids have all the genes of the parents ... the kids are still different. And we know this is true because egg production went up. This couldn't have happened unless there was something (crossover or mutation) that changed from generation to generation. Regarding James'
Re: [FRIAM] Real-world genetic algorithm example... help!
John, Thanks. I agree. In fact, I would argue that ANY attempt to squeeze spiritual juice from this particular example blunts it scientific edge. To mix a metaphor. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] [Original Message] From: John Kennison jkenni...@clarku.edu To: russ.abb...@gmail.com russ.abb...@gmail.com; The Friday MorningApplied Complexity Coffee Group friam@redfish.com Date: 7/10/2010 4:02:16 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! Selecting for productive coops rather than productive hens might reject highly productive, highly aggressive hens in favor of somewhat less productive, considerably less aggressive hens who would leave their coop-mates in peace (and therefore able to produce more eggs). Such hens need not have anything like a concept of loyalty to the coop --we could redistribute these hens to different coops and without affecting coop productivity. But after a while, we might find we are selecting for highly productive, potentially highly aggressive hens who are strongly inhibited against bothering a hen they grew up with. Then if we redistributed the hens among different coops, coop productivity would decrease. Are there applications to genetic algorithms? It shows you have to be careful about dividing the task to be done into subtasks. You dont want to overlook an algorithm for doing one subtask that provides useful byproducts for another subtask. Instead of selecting for each subtask separately, you might select for teams of algorithms that do the whole task. From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 2:50 AM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Real-world genetic algorithm example... help! It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.commailto:teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story - beyond a nice, illustrative example of how a GA works - is the need to properly define a fitness function. In the case of individual chickens, the fitness function was ill-defined and didn't work very well. In particular, this section points out that it is not necessary to know why a good solution is good. Why doesn't have to come into it ... the fitness function simply ensures that the best solution, no matter what the reasons are for being the best, can emerge from this process. In regards to Russ' complaints, I'm not sure I can agree that no crossover/mutation occurred. I haven't read the original paper yet, just the Huff Post treatment, so I didn't realize that the chicken clusters weren't mixed. That is, I just
Re: [FRIAM] Real-world genetic algorithm example... help!
I have been here before. This is the point in the conversation where Roger Critchlow explains to me what the hell is going or or, i die. Roger? Is there a confusion here concerning what is the analogue of the individual in the genetic algorithm? Nick Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] - Original Message - From: Ted Carmichael To: The Friday Morning Applied Complexity Coffee Group Sent: 7/10/2010 7:11:35 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! Well, in regards to (1), yes, I would guess elitism + mutation is a good description. However, I believe that is enough to qualify as a GA. As I recall, some GA practitioners believe mutation is best, some believe crossover is best, and some feel you should have both, or decide based on the problem. I would guess that it is a minority viewpoint to claim that mutation by itself is not enough to be a GA. For (3), you say that the genetic operators are not explicit, in the chicken coop example. My understanding is that sometimes the genetic operator itself is a part of the genotype, and is thus subject to mutation/crossover as well. In such a case, it wouldn't really be explicit because it would vary across the population. Of course, the final genetic operator is discoverable - i.e., it is recorded in the final solution ... but this doesn't really matter to the programmer. He has no knowledge a priori of what that operator will turn out to be; and further, the final genetic operator is not the point of the GA. Finding a good solution is the desired outcome - the operator is secondary at best. For (2), you correctly imply that a phenotype - or genotype - must be explicitly defined in a computer. Well, sure. The computer is deterministic, and so this information will be recorded somewhere as part of the code of the solution. What I find interesting is this idea that the programmer has to know and care and understand the final solution. I don't think that is the case. Oh sure, in some instances the final solution is quite clear. For example, the TSP will end up with a list of city-pairs that is easily understood. But I can also imagine instances of a GA, say applied to computer code, or a mathematical formula, that becomes so immensely complex that the researcher does not understand why the final solution works. And, as pointed out above, he doesn't have to understand why it works. That it does work well is good enough, isn't it? I just don't see any functional distinction between not caring why the final solution works and - in the case of chickens - not being able to precisely describe how it works or what it looks like. It's kind of like that DARPA funded robot pack-animal ... they don't care, really, what all the final rules are, as long as the robot can walk. In fact, I would suspect that some flexibility in the 'final solution' is allowed, and that the machine learning process is continuously running to some degree. But I am enjoying the discussion, so thanks for that. Cheers, -Ted On Sat, Jul 10, 2010 at 2:50 AM, Russ Abbott russ.abb...@gmail.com wrote: It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment
[FRIAM] Virtual-world genetic algorithm example... help!
I am reminded of two conflicting reports I got from two friends about an attempt to evolve a sorting program. One friend reported that it was discouraging. The evolved programs never were reliable and they took all kinds of time and had many superfluous features. The only way to actually get an algorithm that worked was to have a sorting method in mind then give the program more survival credit the more it mimicked the program in mind. Another friend reported that the attempt was a phenomenal success. A program evolved which sorted lists perfectly and efficiently and which was unlike any known sorting algorithm, In fact, no on could figure out what the program was doing; the only reason they felt it most be theoretically correct was that it sorted a huge number of lists perfectly every time. Can any of you tell me which friend is giving a more accurate account? (It is possible that the accounts refer to different experiments and are both accurate. The pessimistic account was told to me about 10 years ago, the other account recently.) FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Virtual-world genetic algorithm example... help!
I've had both experiences. The successful version had a couple of advantages. It had more useful primitives and a more useful fitness function. I don't remember the details, but a primitive that says swap adjacent cells if one is less that the other helps a lot! A fitness function that counts the number of elements out of place is much less useful than one that measures the extent to which the result is ordered, e.g., how many elements are on the correct side of their neighbors. The bottom line is that there has to be a path from the initial primitives to the goal in which each step has increasing fitness. If you've got that an evolutionary process should get there. If not, it probably won't. -- Russ On Sat, Jul 10, 2010 at 1:22 PM, John Kennison jkenni...@clarku.edu wrote: I am reminded of two conflicting reports I got from two friends about an attempt to evolve a sorting program. One friend reported that it was discouraging. The evolved programs never were reliable and they took all kinds of time and had many superfluous features. The only way to actually get an algorithm that worked was to have a sorting method in mind then give the program more survival credit the more it mimicked the program in mind. Another friend reported that the attempt was a phenomenal success. A program evolved which sorted lists perfectly and efficiently and which was unlike any known sorting algorithm, In fact, no on could figure out what the program was doing; the only reason they felt it most be theoretically correct was that it sorted a huge number of lists perfectly every time. Can any of you tell me which friend is giving a more accurate account? (It is possible that the accounts refer to different experiments and are both accurate. The pessimistic account was told to me about 10 years ago, the other account recently.) FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Virtual-world genetic algorithm example... help!
Thanks Russ. depending on the primitives chosen, this could be more in line with the pessimistic account. Putting in the swapping primitive seems like aiming for the simple sort which keeps on swapping until it can't be done anymore. Do you know of any evolutionary process which produced a highly efficient and previously unknown sorting algorithm? ---John From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 8:32 PM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Virtual-world genetic algorithm example... help! I've had both experiences. The successful version had a couple of advantages. It had more useful primitives and a more useful fitness function. I don't remember the details, but a primitive that says swap adjacent cells if one is less that the other helps a lot! A fitness function that counts the number of elements out of place is much less useful than one that measures the extent to which the result is ordered, e.g., how many elements are on the correct side of their neighbors. The bottom line is that there has to be a path from the initial primitives to the goal in which each step has increasing fitness. If you've got that an evolutionary process should get there. If not, it probably won't. -- Russ On Sat, Jul 10, 2010 at 1:22 PM, John Kennison jkenni...@clarku.edumailto:jkenni...@clarku.edu wrote: I am reminded of two conflicting reports I got from two friends about an attempt to evolve a sorting program. One friend reported that it was discouraging. The evolved programs never were reliable and they took all kinds of time and had many superfluous features. The only way to actually get an algorithm that worked was to have a sorting method in mind then give the program more survival credit the more it mimicked the program in mind. Another friend reported that the attempt was a phenomenal success. A program evolved which sorted lists perfectly and efficiently and which was unlike any known sorting algorithm, In fact, no on could figure out what the program was doing; the only reason they felt it most be theoretically correct was that it sorted a huge number of lists perfectly every time. Can any of you tell me which friend is giving a more accurate account? (It is possible that the accounts refer to different experiments and are both accurate. The pessimistic account was told to me about 10 years ago, the other account recently.) FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Real-world genetic algorithm example... help!
How is selective breeding / clustering to optimise particular traits in chickens any different from endogamous human clusters / societies? In India for eg. the endgamous caste and sub-caste systems have been in place for millenia to ensure genetic optimisation and perpetuation of a few desirable traits. My mother will be comforted to learn this has been confirmed by experiments on chickens. Previously all Bengali Brahmins had to rely on were encyclopedias / papers like this [1http://en.wikipedia.org/wiki/Haplogroup_R1a1_%28Y-DNA%29] to confirm that we are bred to perpetuate an R1a1 gene. rol Sarbajit On Sat, Jul 10, 2010 at 8:31 PM, Nicholas Thompson nickthomp...@earthlink.net wrote: John, Thanks. I agree. In fact, I would argue that ANY attempt to squeeze spiritual juice from this particular example blunts it scientific edge. To mix a metaphor. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] [Original Message] From: John Kennison jkenni...@clarku.edu To: russ.abb...@gmail.com russ.abb...@gmail.com; The Friday MorningApplied Complexity Coffee Group friam@redfish.com Date: 7/10/2010 4:02:16 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! Selecting for productive coops rather than productive hens might reject highly productive, highly aggressive hens in favor of somewhat less productive, considerably less aggressive hens who would leave their coop-mates in peace (and therefore able to produce more eggs). Such hens need not have anything like a “concept” of loyalty to the coop --we could redistribute these hens to different coops and without affecting coop productivity. But after a while, we might find we are selecting for highly productive, potentially highly aggressive hens who are strongly inhibited against bothering a hen they grew up with. Then if we redistributed the hens among different coops, coop productivity would decrease. Are there applications to genetic algorithms? It shows you have to be careful about dividing the task to be done into subtasks. You don’t want to overlook an algorithm for doing one subtask that provides useful byproducts for another subtask. Instead of selecting for each subtask separately, you might select for teams of algorithms that do the whole task. From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 2:50 AM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Real-world genetic algorithm example... help! It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as you could push it. If I were forced to describe this in GA terms, I would say that the coop genome is the sequence, in some arbitrary order, of chicken genomes. To get an offspring, take a coop genome and treat the segments that correspond to individual chickens as separate genomes, mate them to get offspring, and then concatenate the genomes of the resulting offspring to get a new coop genome. I've never heard of a genetic operator like that, but I guess that doesn't mean you couldn't claim it as a genetic operator. The bottom line for me though is that the experiment is great biology, but it's a pretty limited and confusing example of a GA. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Fri, Jul 9, 2010 at 11:12 PM, Ted Carmichael teds...@gmail.commailto:teds...@gmail.com wrote: Ha! I knew someone would complain about that. First of all, Eric is correct: the main point of the story -
Re: [FRIAM] Real-world genetic algorithm example... help!
Perhaps if I understood the computer side of this conversation better I wouldn't have the feeling that the chicken example is being misunderstood. But I dont and I do (respectively). It should be remembered that no chickens were selected during the conduct of this experiment; only crates. What determined if crates were allowed to contribute to the next generation was the number of eggs that the crate laid. Chickens changed, but selection was for crate egg production. Changed chicken behavior mediated the change in crate reproductive output. Eliot Sober makes an interesting distinction between selection of and selection for. The experiment resulted in the selction of nice chickens, but selection was for crate egg production. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] - Original Message - From: sarbajit roy To: nickthomp...@earthlink.net;The Friday Morning Applied Complexity Coffee Group Sent: 7/10/2010 10:51:22 PM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! How is selective breeding / clustering to optimise particular traits in chickens any different from endogamous human clusters / societies? In India for eg. the endgamous caste and sub-caste systems have been in place for millenia to ensure genetic optimisation and perpetuation of a few desirable traits. My mother will be comforted to learn this has been confirmed by experiments on chickens. Previously all Bengali Brahmins had to rely on were encyclopedias / papers like this [1] to confirm that we are bred to perpetuate an R1a1 gene. rol Sarbajit On Sat, Jul 10, 2010 at 8:31 PM, Nicholas Thompson nickthomp...@earthlink.net wrote: John, Thanks. I agree. In fact, I would argue that ANY attempt to squeeze spiritual juice from this particular example blunts it scientific edge. To mix a metaphor. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] [Original Message] From: John Kennison jkenni...@clarku.edu To: russ.abb...@gmail.com russ.abb...@gmail.com; The Friday MorningApplied Complexity Coffee Group friam@redfish.com Date: 7/10/2010 4:02:16 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! Selecting for productive coops rather than productive hens might reject highly productive, highly aggressive hens in favor of somewhat less productive, considerably less aggressive hens who would leave their coop-mates in peace (and therefore able to produce more eggs). Such hens need not have anything like a concept of loyalty to the coop --we could redistribute these hens to different coops and without affecting coop productivity. But after a while, we might find we are selecting for highly productive, potentially highly aggressive hens who are strongly inhibited against bothering a hen they grew up with. Then if we redistributed the hens among different coops, coop productivity would decrease. Are there applications to genetic algorithms? It shows you have to be careful about dividing the task to be done into subtasks. You dont want to overlook an algorithm for doing one subtask that provides useful byproducts for another subtask. Instead of selecting for each subtask separately, you might select for teams of algorithms that do the whole task. From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 2:50 AM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Real-world genetic algorithm example... help! It's not a good example as an illustration of GA because (1) the selection mechanism to move from one generation to the next is essentially select the best and shake it up. At best you might call that elitism plus mutation. But it is not representative of GA. (2) it has no explicit representation of the genome (3) there are no explicit genetic operators applied to one or more parents to produce children. The issue of whether there is mutation points out that there is no coop genome that is being evolved. Since there is no coop genome, it's hard to say that there is or is not mutation. It certainly isn't a good illustration of mutation for a textbook. You might make the case that the coop genome is the collection of the chicken genomes and that the offspring coop genome is generated from the parent coop genome by breeding the chickens. I guess you could call that mutation of the coop genome.So the mutation operator on the parent coop genome is to breed the chickens to get a new coop genome. But I think that's about as far as
Re: [FRIAM] Virtual-world genetic algorithm example... help!
Russ Abbott wrote: In a system like this though, you always have to start with some primitives. It's really matter of where you can get from the primitives and whether there is a steadily uphill (in terms of fitness) path for getting there. That's a question of how diversity is maintained in population and what kind of transformations are made to the population of programs. If transformations are modular or there is no mechanism for maintaining diversity, then a rugged fitness landscape may well cause problems -- the population can reduce to, in-effect, one individual and be stuck in a rut forever. It's a problem with optimization algorithms in general, not just genetic programming. It's not that one can't include a looping structure as a primitive. It's that GP is not good at using it. I suspect enhanced evaluation mechanisms are needed to influence fitness. I speculate that historical human imperative programing habits aren't particularly helpful either for automated programming (better to have lambdas bound to names and recursion). The size of the expression tree has been used in GP for a long time to encourage parsimonious solutions to be found, but I suspect there hasn't been much work has been done to provide a cost of a calculation. By that I mean stuff like L3 cache misses (how irregular is the memory access pattern?), the maximum depth of the stack pointer (is it non-terminating recursion?), instructions retired (logically how efficient is the calculation?), and total joules used (what does it really take to make CPUs do it?). Optimizing over that space is what quantifies the difference between good and bad programs.. Marcus FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Real-world genetic algorithm example... help!
Exactly. Although this was not (as far as I know) part of the experiment, one could imagine a similar experiment on groups with more structure, e.g., baseball teams. It's the team that wins the most games (or the most important games) that reproduces. That team probably has pretty good players at each position, but almost certainly it has a good team structure and team organization. In other words, they work well together. That's what matters. -- Russ Abbott __ Professor, Computer Science California State University, Los Angeles cell: 310-621-3805 blog: http://russabbott.blogspot.com/ vita: http://sites.google.com/site/russabbott/ __ On Sat, Jul 10, 2010 at 8:17 PM, Nicholas Thompson nickthomp...@earthlink.net wrote: Perhaps if I understood the computer side of this conversation better I wouldn't have the feeling that the chicken example is being misunderstood. But I dont and I do (respectively). It should be remembered that no chickens were selected during the conduct of this experiment; only crates. What determined if crates were allowed to contribute to the next generation was the number of eggs that the crate laid. Chickens changed, but selection was for crate egg production. Changed chicken behavior mediated the change in crate reproductive output. Eliot Sober makes an interesting distinction between selection of and selection for. The experiment resulted in the selction of nice chickens, but selection was for crate egg production. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/http://home.earthlink.net/%7Enickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] - Original Message - *From:* sarbajit roy sroy...@gmail.com *To: *nickthomp...@earthlink.net;The Friday Morning Applied Complexity Coffee Group friam@redfish.com *Sent:* 7/10/2010 10:51:22 PM *Subject:* Re: [FRIAM] Real-world genetic algorithm example... help! How is selective breeding / clustering to optimise particular traits in chickens any different from endogamous human clusters / societies? In India for eg. the endgamous caste and sub-caste systems have been in place for millenia to ensure genetic optimisation and perpetuation of a few desirable traits. My mother will be comforted to learn this has been confirmed by experiments on chickens. Previously all Bengali Brahmins had to rely on were encyclopedias / papers like this [1http://en.wikipedia.org/wiki/Haplogroup_R1a1_%28Y-DNA%29] to confirm that we are bred to perpetuate an R1a1 gene. rol Sarbajit On Sat, Jul 10, 2010 at 8:31 PM, Nicholas Thompson nickthomp...@earthlink.net wrote: John, Thanks. I agree. In fact, I would argue that ANY attempt to squeeze spiritual juice from this particular example blunts it scientific edge. To mix a metaphor. N Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University (nthomp...@clarku.edu) http://home.earthlink.net/~nickthompson/naturaldesigns/http://home.earthlink.net/%7Enickthompson/naturaldesigns/ http://www.cusf.org [City University of Santa Fe] [Original Message] From: John Kennison jkenni...@clarku.edu To: russ.abb...@gmail.com russ.abb...@gmail.com; The Friday MorningApplied Complexity Coffee Group friam@redfish.com Date: 7/10/2010 4:02:16 AM Subject: Re: [FRIAM] Real-world genetic algorithm example... help! Selecting for productive coops rather than productive hens might reject highly productive, highly aggressive hens in favor of somewhat less productive, considerably less aggressive hens who would leave their coop-mates in peace (and therefore able to produce more eggs). Such hens need not have anything like a “concept” of loyalty to the coop --we could redistribute these hens to different coops and without affecting coop productivity. But after a while, we might find we are selecting for highly productive, potentially highly aggressive hens who are strongly inhibited against bothering a hen they grew up with. Then if we redistributed the hens among different coops, coop productivity would decrease. Are there applications to genetic algorithms? It shows you have to be careful about dividing the task to be done into subtasks. You don’t want to overlook an algorithm for doing one subtask that provides useful byproducts for another subtask. Instead of selecting for each subtask separately, you might select for teams of algorithms that do the whole task. From: friam-boun...@redfish.com [friam-boun...@redfish.com] On Behalf Of Russ Abbott [russ.abb...@gmail.com] Sent: Saturday, July 10, 2010 2:50 AM To: The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM]
[FRIAM] Hywel White et al re 2010 and 1995 neutrino mass findings at Los Alamos Neutrino Detector: Rich Murray 2010.07.10
Hywel White et al re 2010 and1995 neutrino mass findings at Los Alamos Neutrino Detector: Rich Murray 2010.07.10 I have been privileged for over 2 years to warmly appreciate many explorers with sophisticated views at Friday Morning Group. As a layman in all areas, I notice that science always expands and evolves, presenting increasingly subtle paradigms that express aspects of the evolving infinite unity that is our shared reality. Physics has accomplished miracles with the paradigm of nested vibrating geometric processes, invoking ever more abstract layers. The current results use over 4 times the amount of mineral oil as in 1995 -- doubling every 7 years. That was the year when the exponential evolution of the Net showed up as a reality for many citizens -- a history accelerating mutation that was initiated in the global physics lab CERN in 1990. Ipso facto, disruption of established social patterns, chaotic arising of multiple networks of human harmonization. Not disutopia, but Golden Age? Neutrinos, the ubiquitous daughters of the weak interaction, start their universe-traversing lives as one of three varieties: ve, vu, or vt. However, like ghosts with an identity crisis, these phantasmal particles find themselves constantly morphing from one variety to another, or oscillating, as they propagate on their long journeys. Great Google! http://www.symmetrymagazine.org/breaking/2010/06/18/miniboone-results-suggest-antineutrinos-act-differently/ Symmetry Breaking blog archive, extra dimensions of particle physics, a joint FermiLab/SLAC publication Neutrinos, the ubiquitous daughters of the weak interaction, start their universe-traversing lives as one of three varieties: ve, vu, or vt. However, like ghosts with an identity crisis, these phantasmal particles find themselves constantly morphing from one variety to another, or oscillating, as they propagate on their long journeys. Now the MiniBooNE experiment has found that antineutrinos, which should follow the same rules as neutrinos, might oscillate in a slightly different way. The results seem to favor a much-debated antineutrino result obtained by the Liquid Scintillator Neutrino Detector experiment in 1990. The MiniBooNE experiment studies these oscillations by creating intense beams of muon neutrinos and antineutrinos, and directing them at an 800-ton sphere filled with mineral oil and located a half a kilometer away from the beam's source. The vast majority of these particles pass through the detector unscathed; however, a few unlucky voyagers pass too close to a carbon nucleus. The neutrinos, or antineutrinos, interact with carbon nuclei, giving scientists a glimpse of the particles' true identities. MiniBooNE counts how many muon antineutrinos oscillate into electron antineutrinos over a relatively short distance. A 1990 result from the LSND experiment at Los Alamos, which used a beam of muon antineutrinos, reported electron antineutrinos appearing about 0.25 percent of the time. The result is difficult for scientists to reconcile in a world with only three active neutrinos. Earlier this week, after nearly three years of running in antineutrino mode, MiniBooNE collaborators announced that they had obtained a result consistent with the findings from LSND. In fact, analyzing the data in the context of a standard two neutrino mixing model favors an LSND-like signal at a 99.4 percent confidence level. However, model-independent tests show there is still a three percent chance that background fluctutations could mimic the data. While this new result is intriguing, a confirmation of LSND will require more data. Interpretations of the latest MiniBooNE results are complicated due to an apparent difference between the way neutrinos and antineutrinos behave. In a prior analysis based on four years of running with a beam of muon neutrinos, the MiniBooNE experiment did not observe significant evidence for muon neutrinos oscillating to electron neutrinos in the energy range expected under the simplest models for explaining the LSND result. However, an excess was observed at lower neutrino energies (below 475 MeV) at a 3 sigma significance that remains unexplained. Interestingly, the MINOS results announced earlier this week also raises the question as to whether neutrinos and antineutrinos behave differently. The MiniBooNE experiment continues to acquire data, and scientists on the project are hoping to nearly double the antineutrino statistics before the experiment finishes acquiring data within the next two years. Future experiments, such as MicroBooNE or BooNE, a proposal to build a second MiniBooNE detector at a near location, could help to shed more light on these results. This story first appeared in Fermilab Today on June 18, 2010. Rhianna Wisniewski http://www.physicsresearch.tk/2010/02/page/3 This is a Physics News Update distributed by Phillip Schewe of AIP Public