Staged_predict for Boosted Tree classifier #51834

<emThe issue related to the performance of the BoostedTree classifier in terms of accuracy.

System information

  • OS Platform and Distribution: Linux Ubuntu 16.04, Windows 10
  • TensorFlow installed from (source or binary): pip install TensorFlow
  • TensorFlow version (use command below): 2.4.1
  • Python version: 3.6

** Describe the model**
BoostedTree classifier is a model introduced by N Ponomareva, T Colthurst, G Hendry.2017.
It is built over the Xgboost idea and learning is done through building one layer of decision tree regressor over N boosting iteration.

Describe the current behavior
Return the accuracy of the ensemble model with n_tree

Describe the expected behavior
I need to have the accuracy of each individual tree in the ensemble model.
For example, the Gradient Boosting of Sklearn has the staged_predict method.
Is there any similar method in tf.BoostedTree as well?

Standalone code to reproduce the issue
You may find an example of the model, here