Flag EDA artifacts with a user-supplied SVM-style model — flag_gazepoint_artifacts

Applies a user-supplied model or prediction function to segment-level artifact features. No pretrained model is bundled, so this function avoids pretending to reproduce any proprietary or externally trained classifier.

Usage

flag_gazepoint_artifacts_svm(
  x,
  model = NULL,
  feature_cols = NULL,
  probability_threshold = 0.5,
  ...
)

Arguments

x: Either raw EDA data or output from prepare_gazepoint_artifact_svm_features().
model: Optional model object or function. If NULL, features are returned with missing artifact predictions.
feature_cols: Optional feature columns used by the model.
probability_threshold: Threshold for artifact probability.
...: Passed to prepare_gazepoint_artifact_svm_features() when x is raw data.

Value

A data frame with artifact probabilities/classes where available.