Skip to content

Training continuation #8

@leffff

Description

@leffff

Add the ability to train pre-fit ensemble on new data

  1. Receive the init_model of StackBoostRegressor or StackBoostClassifier type
  2. Multiply the first ensemble prediction of init_model by number of samples in the initial dataset
  3. Calculate the mean between the number in step 2 and the new target array
  4. Walk down every thee in the current ensemble and continuously update tree leaves

  1. Add more trees if needed

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions