I would like to ask about optimizing SVM classifier using Genetic Algorithm. I have a process in rapidminer like the attached XML but I'm not sure whether what I am doing is right or wrong.
Based on excel, I observe that you are selecting features based on genetic algorithm and train and test them using SVM. If this is correct there is an efficient way to do it that provides stable and less biased result. When you apply feature selection techniques on the whole dataset there might be bias in your learning, so it is better to use feature selection on training data only. If you want to optimize an algorithm, then you need to try different hyperparameters for that algorithm using optimize parameter (evolutionary) operator.
I attached a text file with XML code go through this and inform if you have any questions.
Based on excel, I observe that you are selecting features based on genetic algorithm and train and test them using SVM. If this is correct there is an efficient way to do it that provides stable and less biased result. When you apply feature selection techniques on the whole dataset there might be bias in your learning, so it is better to use feature selection on training data only. If you want to optimize an algorithm, then you need to try different hyperparameters for that algorithm using optimize parameter (evolutionary) operator.
I attached a text file with XML code go through this and inform if you have any questions.
Based on excel, I observe that you are selecting features based on genetic algorithm and train and test them using SVM. If this is correct there is an efficient way to do it that provides stable and less biased result. When you apply feature selection techniques on the whole dataset there might be bias in your learning, so it is better to use feature selection on training data only. If you want to optimize an algorithm, then you need to try different hyperparameters for that algorithm using optimize parameter (evolutionary) operator.
I attached a text file with XML code go through this and inform if you have any questions.