MATLAB Answers


Best Neural Network for Multivariate Regression

Asked by Federico on 16 Jan 2019
Latest activity Answered by Greg Heath
on 16 Jan 2019
Dear all,
I have a series of input environmental variables (such as time of day, temperature, etcetera) that I would like to use to predict a single target variable. I am not sure if said variables contribute linearly to the target variable and therefore I would like to employ a neural network instead than a standard multivariate linear regression model (e.g.: mvregress), but I am lost in the panoply of options of the neural network toolbox in terms of networks, layers, etcetera.
What would be the best options to try?
Thanks and Best Regards


Sign in to comment.




1 Answer

Answer by Greg Heath
on 16 Jan 2019
 Accepted Answer

Accept all parameter settings.
Train 10 models with different random initial weights
If unsuccessful. Repeat with a range of values for number of hidden nodes.
I have posted ZILLIONS of examples in BOTH comp.soft-sys.matlab and answers.
Hope this helps.
Thank you for formally accepting my advice


Sign in to comment.