How to choose neural network model without having data?
Answers
Answered by
0
hii..
that was nice question..
Hello,
I think the best to try first is SVM ! Forget MLP , it is "has been" now for the majority of problems like yours (OF COURSE not for very complex problems with huge amount of data, HERE the new revolution is DeepLearning, which use MLP)
This is simpler because you do not have to define a priori the structure of the net as you mentionned (layers, units ...), and the weight initialisation is not sensitive.
so if u agree with this answer then mark me as brainlist or follow me..
that was nice question..
Hello,
I think the best to try first is SVM ! Forget MLP , it is "has been" now for the majority of problems like yours (OF COURSE not for very complex problems with huge amount of data, HERE the new revolution is DeepLearning, which use MLP)
This is simpler because you do not have to define a priori the structure of the net as you mentionned (layers, units ...), and the weight initialisation is not sensitive.
so if u agree with this answer then mark me as brainlist or follow me..
Similar questions