Filter löschen
Filter löschen

How to solve the “out of memory” problem in Logistic Regression achieved by “glmfit”

2 Ansichten (letzte 30 Tage)
I use “glmfit” to build Logistic Regression. But the training dataset is too huge. I cannot load all the data into memory. Is there any method to build Logistic Regression in Matlab using batch learning? In other words, it can load the training data gradually when training the model.
Thank you so much.

Antworten (1)

Ilya
Ilya am 2 Nov. 2016
Logistic regression on tall (out-of-memory) arrays is supported in 16b through the fitglm function.

Community Treasure Hunt

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

Start Hunting!

Translated by