Can anyone help me with these learning methods in details, please? I need one better learning method which will be fast, less time consuming and flexible.
Dear Amirul! You need to choose the best method for your application. Please read some articles in the application.
Vladimir
Try to start from fuzzy which gives you the base for NN and SVM..
Read "FUZZY SETS AND FUZZY LOGIC- Theory and Applications" by George J.Klir and Bo yuan Book for fuzzy.. Do not stick on time that will not helpful for you (to do research)..
Dear Amirul! You need to choose the best method for your application. Please read some articles in the application.
Vladimir
Hello, Amirul. OK, localization. But what do you want the system to learn? How is it suppose to work?
It is not possible to give you a good suggestion without further information about your system. For example, Reinforcement Learning might be a better choice depending on your application.
If you have enough data, you can try NN or SVM. If you only have few data, fuzzy inference is better. For localization learning, I prefer neighbor based fuzzy inference.
I think you are talking about supervised learning method. So we could not say which one is better method, it depends on how you trained the dataset.
I suggest you to apply SVM and NN both and compare it. It depends how well you train your model and size of dataset.
SVM and LS-SVM fast and accurate
in Classification and prediction
Applying these methods merely in softwares is not a big deal, but to digest them properly tests your mettle.
In my view, you can easily have a look at the application part of all three initially before begining with learning any one in detail. Then, you may proceed with the one which suits your application or study.
Moreover, working out with these techniques in software can be learnt from number of videos available on Youtube. Apart from this, attending at least a week long workshop on Artificial Intelligence can give you a good start to begin research in this area.
With Regards
A S Kedia
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering