Evolutionary computation is a subfield computational intelligence that involves combinatorial optimization problems. Evolutionary computation uses iterative progress, such as growth or development in a population. This population is then selected in a guided random search using parallel processing to achieve the desired end. Such processes are often inspired by biological mechanisms of evolution.

• Farid Bekran

  Combinatorial optimization by bio-inspired algorithms (evolutionary algorithms)

  I'm looking for a research topic or a hot problem in combinatorial optimization. Can anybody recommend a problem? I am especially interested in MST (minimum spanning trees) problems. Thanks.

  Recent replies ⋅ Show All (9)

  • 
    Hussein Abul-Kasim replied

    http://www.iajet.org/iajet_files/vol.2/no.1/Data%20Hiding%20by%20LSB%20Substitution%20Using%20Genetic%20Optimal%20Key-Permutation.pdf

    Mar 28, 2012
• Karthik Kandasamy

  BEAST and MEGA result interpretation

  I am using MEGA and Beast softwares and I need to infer the results obtained from MEGA and BEAST based on Posterior probability, Bootstrap value, Branch length, Standard deviation, Covarience andI am using MEGA and Beast softwares and I need to infer the results obtained from MEGA and BEAST based on Posterior probability, Bootstrap value, Branch length, Standard deviation, Covarience and other parameter given by tracer. Any advice?
• Karthik Kandasamy

  Insilico pathway reconstruction

  I used a cell designer for pathway drawing. To analyse the path, I tried (but was unable) to import the .xml file into cytoscape. I am looking for troubleshooting advice or recommendations for toolsI used a cell designer for pathway drawing. To analyse the path, I tried (but was unable) to import the .xml file into cytoscape. I am looking for troubleshooting advice or recommendations for tools for metabolic network analysis. Thank you.
• Juergen Bock

  Population Size in Particle Swarm Optimisation

  Hello everybody, can anybody point me to a piece of literature that investigates how population size correlates to convergence in PSO algorithms. To what extent does an increase in the populationHello everybody, can anybody point me to a piece of literature that investigates how population size correlates to convergence in PSO algorithms. To what extent does an increase in the population size decrease the number of required iterations to find the optimum? Can the factors the convergence is influenced by be specified generally? Thanks for any hints, Juergen

  Recent replies ⋅ Show All (3)

  • 

    Comment protected

    Join ResearchGate now to read this comment.
• Janaki Wanigasooriya

  Random weighted genetic algorithm

  There is one method for fitness assigning in multi objective GA is named as Random Weigted Genetic algorithm(RWGA).In that algorithm it uses additional external population(say E) and integer varibleThere is one method for fitness assigning in multi objective GA is named as Random Weigted Genetic algorithm(RWGA).In that algorithm it uses additional external population(say E) and integer varible to keep how many solutions to immigrating from E to next generation. How these parameters integrate to solve the next generation? How to update E after finding nect offsprings? How to find non dominated solutions? Hope anybody help me.I am confused. and really need help. Thanks

  Recent replies ⋅ Show All (5)

  • 
    Graeme Smith replied

    Well, I may not be the best person to discuss this with, but let's see if I can pick your problem apart into pieces I can try to help you with. First of all, in a multiple objective case you wantWell, I may not be the best person to discuss this with, but let's see if I can pick your problem apart into pieces I can try to help you with. First of all, in a multiple objective case you want to test each element for each objective, and get a fitness factor for that element/objective. Then, you might want to have a composite fitness factor for a solution that meets all the objectives to some extent. Call it the compromise solution. Now this means that you have N+1 populations to keep track of where N is the number of objectives you are trying to combine. For each population you are tracking in the larger population, any member that is part of one of the other populations but not the one you are tracking, at that point in the program, is a part of E. As well, just to give the system a little larger variation you might want to create the odd random parameter and feed it in as part of E. Ok, now for each population you could say that the dominant solution is the one with the most members that are within Y standard deviations from the centroid of the set. So really the problem is one of doing statistical analysis to see where the population falls on a scatter plot, or numerical equivalent. The non-dominant solutions will be those with a slightly lower probability than the dominant and thus higher populated solution. Again if you aren't getting any secondary solutions you probably need to perturb the system more, or it is a closed system, and doesn't have multiple solutions. The nE designator might have to do with the objective n in essence talking about the members of all the other sets, that do not combine with the n set, during the analysis. The idea of an E for every objective nE, is I think what that designator is talking about. If you have multiple tests, and multiple subsets of the population that are part of a greater population then clearly you have subsets and some of them must be less dominant in the compromise. If not there is not enough separation in your objectives or your selection or fitness test is not demanding enough. Does this help?

    Jul 7, 2011
• Karthik Kandasamy

  Multiple sequence alignment

  I am doing project on evolutionary relationship and pathway reconstruction of anthocyanin pathway on solanaceae family plants and have collected 13 group of gene sequences (850 seqs.). It was veryI am doing project on evolutionary relationship and pathway reconstruction of anthocyanin pathway on solanaceae family plants and have collected 13 group of gene sequences (850 seqs.). It was very difficult in the multiple sequence alignment step because of more number of sequences. How can I prepare the Multiple Seq. Alignment for phyogen. tree construction? I got an error when using MUSCLE in mega software; I got a result in clustalw, but there are many coloumn of sequence alligned - how can I align them manually? Ho should I select columns for phylogenetic tree construction? Thank you.

  Recent replies ⋅ Show All (2)

  • 
    Karthik Kandasamy replied

    IN SWAMI WORKBENCH muscle algorithm is not available.... for large no. of sequence MUSCLE is suitable... thats y im not using that swami workbench...

    Jun 26, 2011
• Karthik Kandasamy

  Evolutionary analysis

  I am working with solanaceae family plants pigment evolutionary analysis and I collecting sequences from different sources. When I blast against any taxonomy I got similar sequences but my doubt isI am working with solanaceae family plants pigment evolutionary analysis and I collecting sequences from different sources. When I blast against any taxonomy I got similar sequences but my doubt is that I am working with pigment evolution in fruits and flowers. Some sequences are predicted from leaf extracts in EST blast hits. But they show more similarity. What should I do?
• Achmad Basuki

  Genetic programming/cellular automatons

  I am interested in evolutionary computation especially genetic programming and cellular automaton.

  Recent replies ⋅ Show All (6)

  • 
    Graeme Smith replied

    Yeah, when I was doing Mudd, I designed a zero-crossing cryptoalgorythm, no reverse engineering on that one! if you tried to reverse it, you just got divide by zero errors ;) I offered it as aYeah, when I was doing Mudd, I designed a zero-crossing cryptoalgorythm, no reverse engineering on that one! if you tried to reverse it, you just got divide by zero errors ;) I offered it as a "Crypt" command for Macs.

    May 25, 2011
• Ibrahim Tholley

  Robot dancing - adapting robot dance to human preferences

  Hi, As part of my research project, I'm looking into how we can make robots dance and adapt their robotic dance movements to the preferences of humans. The ultimate aim of the research is toHi, As part of my research project, I'm looking into how we can make robots dance and adapt their robotic dance movements to the preferences of humans. The ultimate aim of the research is to develop an overal model for dancing robots as my contribution to make robots more socially acceptable in society. So far, I have developed a program to make robots learn the fundamentals of dance and have now modified the program so that the robot changes and adapts it's movement to human preferences, but currently on the process of doing experiments. I was wondering if any you wouldn't mind taking part in this questionnaire, by simply observering some videos of my dancing robot and answering honestly the related questions. The videos and the questionnaire can be found here: http://www-staff.lboro.ac.uk/~coit2/stage4.html I'll also be interested to know if anyone else is doing work in this area or can share ideas etc. Thanks for your help Kind regards ibs

  Recent replies ⋅ Show All (1)

  • 
    Luis Fabrix replied

    Original Text: (Spanish) Hola, soy nuevo en esta red, y bien este es mi primer post. Me interesa mucho el área del la inteligencia artificial, tanto como, la simulación de sistemas o aspectos delOriginal Text: (Spanish) Hola, soy nuevo en esta red, y bien este es mi primer post. Me interesa mucho el área del la inteligencia artificial, tanto como, la simulación de sistemas o aspectos del universo en el cual estamos incluidos. Me ha pasado algo tan absurdo como gracioso e inclusive perturbador, pase casi 1/2 hora escribiendote un post, y dado que no manejo muy bien el idioma ingles, lo escribí en un traductor online, el cual, para mi disgusto, al tratar de traducir el texto se me ha borrado este por completo. Por esto, al no tener suficiente paciencia como para volver a escribirlo, te doy algunas ideas que me surgieron, para mas detalle de estas, estoy dispuesto a brindarte un análisis mas profundo y detallado sobre tu proyecto. Bueno, algunas pistas: *Para lograr un sistema bueno, tendrás que pensar en como te desemvuelves tú al danzar o bailar. *Notarás que tus moviemientos se 'adaptan' al ritmo de la percusión en mayor medida, pues, en general, esta marca el ritmo de la melodía, como sabrás. *El ritmo tiene, creo, en general, 2 partes, 1- la secuencia o ciclo principal, y 2- el estribillo. *Teniendo en cuenta esto, podrías diseñar un sistema que detecte estos 2 ciclos, principales, con sus tiempos y/o frecuencias. De aqui se me ocurre una idea: -Elige unas cuantas, 20 o 30 melodías, y para cada una de ellas, de forma sincronizada, le marcas al sistema del robot, los movimientos y almacenas, tanto el movimiento, que lo puedes codificar con un simple número, donde este represente algun tipo de movimiento del robot, y como segundo dato, el momento en que le induces tal movimiento, es decir que te queda un vector con registros dobles o de 2 datos. La siguiente etapa, quizas la mas difícil, es diseñar un sistema que, analize las secuencias acústicas, junto a su vector correspondiente, y que pueda determinar el patrón común en todas las secuencias, basicamente que hay en comun entre cada par señal sonora-vector de datos. Si puedes conseguir un sistema que haga esto por tí, bueno, ya tienes la posibilidad de hacer un robor bailarín muy sofisticado. Luego, solo analiza como haces tú, variaciones en tus movimientos, para no caer un algo monótono y aburrido, esta parte quizas es la mas sencilla, y seguro se puede, dadas las condiciones, diseñar un sitema 'experto' en baile, que básicamente se adapte a cualquier melodía, y pueda inclusive, bajo una supervicion que le brinda lo bueno o malo de su baile, aprender a hacer complejos movimientos al ritmo de la musica. Bien, por ahora es lo que se me ha ocurrido, pero como te he comentado, sería para mí, muy interesante, si lo deseas, y sin ningún tipo de compropmiso en principio, colaborar contigo, realizando un análisis mas profundo. Saludos, cordiales, atte. Luis F. --------------------------------------------------------------------------------------------------------------- Online Translation: Hello, I am new in this network, and this good is my first post. I am interested in very much the area of the artificial intelligence, so much like, the simulation of systems or aspects of the universe in which we are included. Something so absurd has happened to me as gracefully and inclusive disturbing, happen almost 1/2 hour writing to you a post, and since I do not handle very well the English language, I wrote it in a translator online, which, for my grief, on having tried to translate the text one has erased this one me completely. That's why, on not having had sufficient patience like write it again, I give you some ideas that me arose, for more detail of these, I am ready to offer to you an analysis deeper and detailed on you project. Good, some tracks: *Para to achieve a good system, you will have to think in like you desemvuelves you on having danced or to dance. *Notarás that your moviemientos 'adapt themselves' to the rhythm of the percussion in major measurement, so, in general, this one marks the rhythm of the melody, since you will know. *El rhythm has, I believe, in general, 2 parts, 1 - the sequence or principal cycle, and 2 - the refrain. *Teniendo in account this, you might design a system that detects these 2 cycles, principal, with his times and/or frequencies. Of here an idea occurs to me: - it chooses a few, 20 or 30 melodies, and for each of them, of in sync form, you mark him to the system of the robot, the movements and you store, so much the movement, that you can codify it with a simple number, where this one represents some type of movement of the robot, and as the second fact, the moment in which you induce such a movement, that is to say that you have a vector left with double records or of 2 information. The following stage, perhaps the most difficult, is to design a system that, analize the acoustic sequences, along with his corresponding vector, and that the common boss could determine in all the sequences, basically that there is together between every pair sonorous sign vector of information. If you can obtain a system that does this for you, good, already you have the possibility of making very sophisticated ballet dancer a robor. Then, only he analyzes as bundles you, changes in your movements, not to fall down anything monotonously and boring, this part perhaps is the simplest, and insurance can, considering the conditions, design to itself a 'expert' sitema in dance, that basically he adapts himself to any melody, and it is possible inclusive, under a supervicion that it drinks to him the good or bad of his dance, to learn to do complex movements to the rhythm of the music. Well, for the time being this is what has occured to me, but since I have commented on you, it would be for me, very interesting, if you wish it, and without any type of compropmiso at first, to collaborate with you, realizing a deeper analysis. Greetings, cordial, atte. Luis F.

    May 20, 2011
• Janaki Wanigasooriya

  GA related Question

  Can any body help me to understand this An over enthusiastic “biology envy” can only be to the detriment of both disciplines by masking the broader potential for two-way intellectual traffic ofCan any body help me to understand this An over enthusiastic “biology envy” can only be to the detriment of both disciplines by masking the broader potential for two-way intellectual traffic of shared insights and analogizing from one another. Three fundamental features of biological evolution illustrate the range of potential intellectual flow between the two communities: particulate genes carry some subtle consequences for biological evolution that have not yet translated mainstream EC; the adaptive properties of the genetic code illustrate how both communities can contribute to a common understanding of appropriate evolutionary abstractions; finally, EC explo- ration of representational language seems pre-adapted to help biologists understand why life evolved a dichotomy of genotype and phenotype

  Recent replies ⋅ Show All (1)

  • 
    Jesús-Antonio Hernández-Riveros replied

    Miss Janaki, I can not precise your question to give you some help. Where is your text extracted from? First Feature: EC is only an aproximation to real world biology, nowadays EC is working withMiss Janaki, I can not precise your question to give you some help. Where is your text extracted from? First Feature: EC is only an aproximation to real world biology, nowadays EC is working with statistical operators that coud be one step to pass from individuals to species. Second Feature: yes, both communities (I think you refer to Biology and EC) contribute to themselves. Third Feature: Original evolutionary algorithms were inspired in biology, the actual tendency is working directly with phenotypes (real vector representation) and mathematical operators. JAHR

    Apr 25, 2011
• Adrian Toader-Williams

  WE need contributions from ALL EXPERTS / PROFESSIONS

  Please Join the MULTIDISCIPLINARY group: Ecology & Economics and Non-Monetary Values. The Role of States and Governments WE need contributions from ALL EXPERTS / PROFESSIONS Please Go to thePlease Join the MULTIDISCIPLINARY group: Ecology & Economics and Non-Monetary Values. The Role of States and Governments WE need contributions from ALL EXPERTS / PROFESSIONS Please Go to the LINK, JOIN IN and place / POST your comments ON: http://www.researchgate.net/group/Ecology_Economics_and_Non-Monetary_Values_The_Role_of_States_and_Governments/ See you on the group! - See perhaps your contributions as well. Thanks!