NNSOM.utils.get_perc_misclassified

NNSOM.utils.get_perc_misclassified(target, prediction, clust)

Compute the percentage of misclassified samples in each cluster.

Parameters:

• target (array-like, shape (N,)) – True class labels.

• prediction (array-like, shape (N,)) – Predicted class labels.

• clust (list of array-like) – A list of cluster index arrays.

Returns:

proportion_misclassified – Percentage (0–100) of misclassified samples per cluster.

Return type:

np.ndarray, shape (numNeurons,)