How to build a multiple output regression model?

The problem I am trying to solve is to build a regression model that maps "n" independent variables to "m" response variables.

I have nearly 35000 data points for each of the "n" independent variables and I want to build a regression model using this 35000 X n space to obtain relations to 35000 corresponding data points to each of the "m" response variables (35000 X m response space). Note: eveyrthing is a "double" data type

I found 'fitrauto" function for hyper parameter optimzation for each of the output variables individually by choosing the best regression model and optimising the corresponsing parameters. But what I would like to know is if there is an equivalent function that can build and optimize a regression model for my multi-input, multi-output case.

Schematically what i would like to do:

table_with_data=table(var1, var2, ..., varn)

regression_model=awesome_function(table_with_data, {response_variables}) %[hopefully a function similar to fitrauto so that i don't manually need to evaluate different regression models]

here {response variables}=set of variables {var_a, var_b,...var_m}

I would like to know if there is such an awesome_function, if not how could I implement one?

Answer 1

[Vimal Rathod]

Hi,

Currently there might not be any function like 'fitrauto' for multi-response variable regression but you could create your custom function using a bayesian model to optimize hyperparameters using the following link.

https://in.mathworks.com/help/deeplearning/ug/deep-learning-using-bayesian-optimization.html#responsive_offcanvas

[Alejandro Plata]

If you don't mind me asking, which MATLAB functions support multi-response regression? I have been searching through mathworks and have been entirely unable to find a suitable function. Thank you!