tensorflow/decision-forests

Faild in multi-class classification

mohammad69h94 opened this issue · 4 comments

In the tensorflow website, You mentioned that "RandomForestModel" support multi-class classification. In composing ANN with RFM we have problem, it's not possible to do this.
Nobody could not have use it as multi-class until now. And then please correct your opinion.

rstz commented

Hi,

Tensorflow Decision Forests supports multi-class classification. Our Beginner tutorial shows how to use TF-DF with Random Forests for multi-class classification on the penguins dataset.

Btw, just in case, notice that multi-class != multi-label (wikipedia).

Hi,

Tensorflow Decision Forests supports multi-class classification. Our Beginner tutorial shows how to use TF-DF with Random Forests for multi-class classification on the penguins dataset.

If you consider my comment, I wrote in composing ANN and decision forest. This not possible to do this. https://www.tensorflow.org/decision_forests/tutorials/model_composition_colab

Btw, just in case, notice that multi-class != multi-label (wikipedia).

If you consider my comment, I wrote in composing ANN and decision forest. This not possible to do this.
https://www.tensorflow.org/decision_forests/tutorials/model_composition_colab