I was wondering if anyone has experience using Matlab Genetic Algorithm Toolbox and could provide help with the coding and such. I am not asking for one to write the code for me but anyone that can provide help if I have questions
Hi,
I have experience with Matlab and with his power to compute such kind of operations. From my opinion OptimizationToolbox doesn’t fullfield my necessity’s, and is a little bit fuzzy so I have used Python or more precise a mix of libraries written in Python.
If you really need Matlab Tolbox I can help you. Thx
Hi, I had some experience on GA Matlab tool. I feel the best way is perhaps to write your own version to suit your needs. In many real applications, the GA is combined with specially designed local search algorithm. In this sense, you might better to do the programming from the scratch.
I have some experience with the GA toolbox, but up to Matlab 2010 version. Do you need the GA toolbox for research in GA or for research only applying GA? For the first, it is better to use libraries in Matlab or another language (Java, C++, Python, etc.) because GA toolbox limits you to some predefined set of encodings, crossover operations, selection methods (only roulette and tournament as far as I remember). But if you use GAs is not the core of your research but it is only used as an optimizer for another research area, Matlab GA toolbox is very easy, you only have to program a matlab function (m file) representing the fitness, with parameters corresponding to the variables you want. If you have questions, drop them here.
Genetic Algorithm belongs to the Global Optimization Toolbox:
http://www.mathworks.co.uk/help/gads/genetic-algorithm.html?searchHighlight=genetic
I do agree that sometimes it is best to customize your own Genetic Algorithm version for your own problems. However, GA in Global Optimization Toolbox provides the framework for this. In other words, you can customize your GA by providing your fitness function, custom coding Selection, Crossover and Mutation functions and setting up other GA parameters. It is therefore a convenient way to create your own customized GA from an existing framework, rather than coding up the whole algorithm from scratch.
You can also easily speed up your GA computation by making use of parallel computing: for that purpose (if you have access to Parallel Computing Toolbox and your MATLAB parallel pool is opened) set GA option 'UseParallel' to 'always'.
I do agree, the best way is write up my own GA codes from scratch instead of using the existing GA toolbox. Its save my time and easier. Thanks for your helps and comments really appreciate it.
Zied, from my experience, yes you can modify the GA tool, however, you cant point Matlab to read the data form outside. Therefore, you have to import the data into Matlab before run the GA tool. However, as I mentioned before, this is based on my experience thus correct me if i'm wrong friends.
I think it is very easy to use GA Toolbox. Just call your objective function, Set constraints and run. But you should take care to select some parameter like
Population options
Population type
Population size etc.
Thank You.
Regards
Suman Debnath
Someone knows how can I do to increase the precision of each gen of the algorithm?. I am using the toolbox, but I just can see the "Bound option" but nothing precision.
I would like to use the number biggest of decimals
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs