How we do regression when output in a matrix?

[dhoa (Dien-Hoa)]

I want to do a regression with input is a matrix and output is a vector. Dimension: Input - 132x32 and output 1024x1. output values is just small integer (0 … 15)

I know how to do the regression when output is a scalar but not sure we have a method when output is a matrix.

I find in the thread below that we can make a network for each variable in output. However I need to reduce the computational complexity as much as possible so I’m not sure this is a good approach.

How to predict multiple regressions from a model?

Maybe a starting approach would be a network for each variable. You would train a model for number of assists, a model for number of blocks and so on. Using a plain MSE as loss function in a “multi-regression” layer could be problematic since the variables are on different scales. Do you have many attributes to predict?

I’m thinking maybe I can do a autoencoder to reduce the dimension. But anyway, i’m new in this topic so there might have a better approach.

I’m very appreciated if someone can help me on this problem. Thank you a lot

[cthnguyen]

Hello ! I’m also trying to predict a matrix as a regression problem. Did you manage to find a way to solve your problem?