hyperparameter_tuning: hyperparameter_tuning

In rosepeglershare/TagR: SHARE's TagR package that uses machine learning to identify a comment's relevant topics

Description

Finds the best set of xgboost parameters for each topic using random search.

Usage

hyperparameter_tuning(
  train_labelled_dtm,
  valid_labelled_dtm,
  train_labels,
  val_labels,
  topics,
  num_its = 1000
)

Arguments

train_labelled_dtm	Training labelled document-term matrix.
valid_labelled_dtm	Validation labelled document-term matrix.
train_labels	Training labels matrix.
val_labels	Validation labels matrix.
topics	List of topics.
num_its	Number of iterations to run for each topic. Default: 1000

Details

Parameters:

• max_depth: Maximum depth of a tree. Increasing this value will make the model more complex and more likely to overfit.

• eta: Step size shrinkage used in update to prevent overfitting.

• subsample: Subsample ratio of the training instances. Setting it to 0.5 means that XGBoost would randomly sample half of the training data prior to growing trees and this will prevent overfitting.

• colsample_bytree: The subsample ratio of columns when constructing each tree. Subsampling occurs once for every tree constructed.

• min_child_weight: Minimum sum of instance weight (hessian) needed in a child. If the tree partition step results in a leaf node with the sum of instance weight less than min_child_weight, then the building process will give up further partitioning.

Value

A dataframe with columns representing parameters and rows representing an optimal parameter set for each topic.

rosepeglershare/TagR documentation built on Dec. 31, 2020, 3:12 a.m.