validation question

[namasang1]

hello, I read your paper interestingly and want to apply it to my custom data.
In this regard, two problems occurred.

first, when i evaluate my custom data, i can suffer this error

do you know what is?

And can i extract each prediction results, not average prediction result?.

thank you

[kaiwenzha]

Hi @namasang1 Thanks for your interest. It seems that the error was raised because the low_shot_gmean list is empty, which indicates that there is no low-shot label under the current low shot threshold. You may change the many/medium/few shot thresholds (check many_shot_thr and few_shot_thr parameters in the following code) according to your data.

imbalanced-regression/agedb-dir/train.py

Line 338 in 055a7b3

def shot_metrics(preds, labels, train_labels, many_shot_thr=100, low_shot_thr=20):

Sure, you can extract each prediction result without averaging. preds in the following code is a list that saves all the prediction results. You can directly use it without applying shot_metrics() function to get averaged results.

imbalanced-regression/agedb-dir/train.py

Line 325 in 055a7b3

shot_dict = shot_metrics(np.hstack(preds), np.hstack(labels), train_labels)