what is the 30 dimensions of predicted result?

[CecileLiu]

The predicted value is a 30-dimension numpy array. What is the meaning of the score of each dimension? It seems not specify in the paper.

[stasaki]

We used a data augmentation approach during the prediction phase to generate 30 variations of an image and average the predictions. This helps ensure that small differences in image quality don't bias the model's predictions. We have included this detail in the revised manuscript that will be published soon. I will also update the document on GitHub. Thanks for bringing this to my attention.

…

On Sun, Aug 13, 2023 at 10:07 AM Cecile ***@***.***> wrote: The predicted value is a 30-dimension numpy array. What is the meaning of the score of each dimension? It seems not specify in the paper. — Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABNASFQIVNLSGY5HI2KTMW3XVDUSPANCNFSM6AAAAAA3OWRQXQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[CecileLiu]

Thank you for the explanation. I have further questions:

1. Is the values in the .txt file of phenotype folder the grounded true?
2. If yes, what does it mean for higher or lower the value?

[stasaki]

All data in the 'data' directory is purely synthetic. It is primarily used to demonstrate the data format. For real data, please visit https://www.radc.rush.edu.

…

On Sun, Aug 13, 2023 at 11:12 AM Cecile ***@***.***> wrote: Thank you for the explanation. I have further questions: 1. Is the values in the .txt file of phenotype folder the grounded true? 2. If yes, what does it mean for higher or lower the value? — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABNASFXJFPD3EKTM25NDMMLXVD4ILANCNFSM6AAAAAA3OWRQXQ> . You are receiving this because you commented.Message ID: ***@***.***>