I am working on a 3 classes classification problem and my confusion matrix is getting messed up. It does not show class 1.
The test accuracy for all three classes is 95% and all three classes are actually in the dataset.
# log data for the confusion matrix
wandb.log({"conf_mat" : wandb.plot.confusion_matrix(probs=None,
y_true=y_test_true, preds=y_test_pred,
class_names=labels)})
Any idea how to solve this issue?
