This feature preprocessor object applies an ANOVA to the training data to find the p-value of all features. It then either uses the top k features with the smallest p-values, or it removes the features with the smallest k p-values. Additionally, this function can be used to remove the top k p-values and then use only the following j next smallest p-values (for example, this can be useful if one is interested in comparing the performance using the most selective 10 neurons to using the next most selective neurons, etc.).
Properties and Methods
The following properties can be set to control the behavior of this feature preprocessor (note that either num_features_to_exclude num_features_to_use must be set)
fp = select_or_exclude_top_k_features_FP
This property sets the number of features with the smallest p-values to exclude. This method or set_num_features_to_use (or both) must be set prior to using the object to normalize data.
This property sets the number of features with the smallest p-values to use. This property or num_features_to_exclude (or both) must be set prior to using the object to normalize data.
fp = save_extra_info (default = 0)
If this property is set to 1, the p-values for all features will be saved and returned along with the classifier results. It should be noted that setting this value to 1 will greatly increase the size of the saved results file.
The following methods are used by the cross-validator algorithm to apply feature preprocessing to the data:
[fp XTr_preprocesed] = set_properties_with_training_data(fp, XTr, YTr)
Calculates the p-values for each feature by applying an ANOVA to the training data (XTr, YTr). The method then returns the training data with either only the top k most selective neurons, or with the top k features removed (or with the top k features removed and only the next j features used, if both set_num_features_to_use and set_num_features_to_remove have been set).
X_preprocessed = preprocess_test_data(fp, X_data)
Removes or only uses the top k selective neurons (as determined using the training data). This method is usually applied by the cross-validator to the test data.
current_FP_info_to_save = get_current_info_to_save(fp)
Returns the p-values from ANOVAs applied to each feature in the structure current_preprocessing_information_to_save.the_p_values_org_order. These values will be saved the cross-validator algorithm.