Generate Predictions or Retrodictions From a Set of (Adaptively) Boosted Trees and Model Weights

This function implements the testing phase of the AdaBoost algorithm. It extracts the adaptively boosted weak learners (e.g. classification stumps) and their corresponding weights to combine them into the weighted sum $$H(\mathbf{x}) = \sum_{i=1}^Ta_th_t(\mathbf x),$$ which yields AdaBoost's predictions or retrodictions on a given set.

Usage

testAda(fit, data, input_checks = TRUE, verbose = TRUE)

Arguments

fit

A model fitted with trainAda. This is a list of boosted trees and theirs weights. More specifically, each element of the list contains a weak learner (e.g., a decision stump) and its corresponding weight.

data

A data frame containing the test set for predictions. The structure should match the training set used to generate h.

input_checks

A logical value indicating whether to perform input validation checks. Defaults to TRUE.

verbose

A logical value specifying whether to display progress messages and animations. Defaults to TRUE.

Value

A numeric vector containing the final predictions from the AdaBoost model. The predictions are in the form of -1 or 1, corresponding to binary classification outcomes.

Details

The function proceeds as follows:

1. If input_checks is TRUE, basic input validation is performed to ensure that fit is a valid AdaBoost model and data is appropriate for predictions.

2. Progress messages and animations are shown if verbose is TRUE.

3. Boosted weak learners (h) and their weights (a) are extracted from fit.

4. Predictions or retrodictions are generated for each weak learner using the test set.

5. Individual predictions are combined using the weights from the AdaBoost model to produce the final predictions.

Examples

# Example usage:
# Assume `fit` is a trained model and `test` is a data frame.
# ypred <- testAda(fit, test)