Outlook for multi-output regression?

Feature requests
#1

Hello everybody,
i wonder if there is anything on the roadmap of edge impulse to help with multi-output regression?
There is this article regarding regression, but it i think this only applies to one output.

The idea i would like to test is if a multi-output model could help in a localization project in which the signal strength of one mobile beacon at multiple fixed sensors can be used to predict the location of the beacon.

The input would be signal strength measurings on different locations (lets say x and z axis) and the two outputs should be the x and z axis for given signals strengths at each sensor in a given moment.

Maybe i am wrong, but this might be a usecase for multi-output regression.

Thanks a lot

#2

Hello @clemens_edge,

We don’t support multi output for regression models out of the box but feel free to try something like that with the BYOM feature: DataTechNotes: Multi-output Regression Example with Keras Sequential Model

Also, I’ll share your request with our ML team.

Best,

Louis