How to specify matrix type response and predictor variables for a regression model

2 views (last 30 days)
I have created a neural network with a regression output layer. My response as well as predictor data are 1000*2 matrices. Currently, I am only able to specify a single data sample to the network when I use trainNetwork, but I am not able to feed the entire dataset. I tried to solve this problem by using 3D arrays, but that is not working. Can anyone please help me out? Currently I am only able to input: trainNetwork(L1, P1, layers, options), and I have matrices ranging from L1 to L50 and similarly for P, and I want to specify all these matrices as response and predictor data respectively.

Accepted Answer

Madhav Thakker
Madhav Thakker on 29 Jun 2021
Hi Sumedh,
I understand you want to train on more than 1 input datapoint at a time. Have a look at batching documentation to see if this helps you.
  1 Comment
Sumedh Soman
Sumedh Soman on 30 Jun 2021
Hello, Thanks a lot for the answer, the approach I followed after reading batch documentation has worked. Thanks for helping me out!

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by