catboost.sum_models {catboost}R Documentation

Sum models.

Description

Blend trees and counters of two or more trained CatBoost models into a new model. Leaf values can be individually weighted for each input model. For example, it may be useful to blend models trained on different validation datasets.

Usage

catboost.sum_models(
  models,
  weights = NULL,
  ctr_merge_policy = "IntersectingCountersAverage"
)

Arguments

models

Models for the summation.

Default value: Required argument

weights

The weights of the models.

Default value: NULL (use weight 1 for every model)

ctr_merge_policy

The counters merging policy. Possible values:

  • 'FailIfCtrIntersects' Ensure that the models have zero intersecting counters

  • 'LeaveMostDiversifiedTable' Use the most diversified counters by the count of unique hash values

  • 'IntersectingCountersAverage' Use the average ctr counter values in the intersecting bins

Default value: 'IntersectingCountersAverage'

Value

Model object.

[Package catboost version 1.0.4 Index]