Which method should I take for regression neural network in Matlab?

Dear all,

I intend to use Matlab for a regression neural network task. I need to provide a full dataset and then train a regression model based on neural network algorithms to predict new valus of the data.

In this regard, as far as I know following methods are doable in Matlab:

1) nntool

2) Regression learner App

2) write code

Which one is suggested for beginners?

Thansk a lot

Answer 1

[Gouri Chennuru]

Hi Alireza,

You can make use of the Regression Learner App available in MATLAB, as it helps to choose, various algorithms to train and validate regression models. After training multiple models, it also compares the validation errors side-by-side, and then chooses the best model.

If you want to implement it using nntool method, instead of nntool method, I recommend you to use nnstart method which provides graphical interfaces that allow you to design and deploy fitting, pattern recognition, clustering, and time-series neural networks.

Hope this Helps!

[Alireza Babaei]

Hi Gouri,

Thanks for your comprehensive help.

Do you know how can I install 'Regression Learner App' in Matlab version 2015? (I don't have it)