Dear all,
I am a PhD student with little experience in machine learning. I have to optimize a function whose analytical expression in unknown, but I have a matlab script which computes its values point by point. This function depends on n discrete variables and on m categorical variables. Which tecniques of machine learning may help me to compute the (sub)optimum points? I was thinking about genetic algorithms, but may also other tecniques (NN, SVM, ...) work properly? If so, how could they be employed in this particular case?
Many thanks for your kind attention!
Best,
Fabio.
you can try trees (cart/rpart)/random forests, they work well for categorical inputs.
Many thanks Vikas. I hadn't thought about random forests! But also ga may work? And how about the other tecniques?
Many thanks for your attention!
Dear Fabio Della Marra
Read the following book may be help you
"Machine Learning: Modeling Data Locally and Globally (Advanced Topics in Science and Technology in China)"
- Badges
- Science topic