-1

I'm using "system.speech" library to recognize voice commands, but I want to use genetic algorithms to improve recognition, and I'm new to GAs and I've some questions:

  • How to implement this (what is the population in this situation? the genomes..)

  • Can this be done using the c# speech recognition library or should I use FFT....

Any other advice ll be apprecited. Thank you

Scott
  • 41
  • 1
  • 6
  • Vote here: http://area51.stackexchange.com/proposals/57719/artificial-intelligence :) – BartoszKP Oct 15 '13 at 13:44
  • But even there: claiming that one language is better than another to implement such a general concept is, in general, pure nonsense. And asking about it is too broad. Choose the language you feel more comfortable with. Second question is also too broad. How should anyone now this beforehand? Take a library, and check if it suits your needs for yourself. If not, take another, or try your own implementation. – BartoszKP Oct 15 '13 at 13:46
  • This is indeed off-topic for StackOverflow because you are asking opinion-based questions. This would probably be off-topic for any StackExchange site. It would be more appropriate for some kind of discussion forum, not a QA site like this. – Gray Oct 15 '13 at 13:46
  • Thank you, I'm new to genetic algorithms and I need help and advice, and I want to use C# to achieve that. I think the right question is: How can I use GA to improve VCR? what are the steps? I didn't say some language is better than another, I read that may be some language are more suitable to do some threads, and also so many people use them to do so, so I'll find more help... Hope you understand What I meant. And sorry – Scott Oct 15 '13 at 14:05

1 Answers1

0

I thinf that a neural net is better for this tasks, because give you directly the functional relation between inputs and output. Your problem is a tipical classification issue with discrete output, I think.

Francisco Yepes Barrera
  • 156
  • 4