I'm trying to use scikit-learn to plot a confusion matrix from raw data I have obtained (contains just predictions and ground truths).
The data contains a total of 4 classes:
[0, 1, 2, 3]
One of the parameters of the confusion matrix is sample weight, and I noticed the shape has to be equal to the number of samples in the data.
Considering that the classes are imbalanced, and given the class ratio to all of:
[0.5, 0.3. 0.15, 0.05], what would I need to pass to
sample_weights to account for the class imbalance in evaluation, when I'm also trying to normalize the matrix?