Zobrazeno 1 - 3
of 3
pro vyhledávání: '"Seyedamir Shobeiri"'
Autor:
Seyedamir Shobeiri
Publikováno v:
Iraqi Journal of Information & Communication Technology, Vol 7, Iss 2 (2024)
This paper intends to provide a better understanding of how these models produce predictions, particularly in complex medical diagnoses, and at the same time bridging the gap between technical model outputs and clinical applications. This study addre
Externí odkaz:
https://doaj.org/article/67d6e87127844e7a8c3f1a3bd6764cf8
Autor:
Seyedamir Shobeiri, Mojtaba Aajami
Publikováno v:
Iraqi Journal for Electrical and Electronic Engineering. 18:9-14
Low-quality data can be dangerous for the machine learning models, especially in crucial situations. Some large-scale datasets have low-quality data and false labels, also, datasets with images type probably have artifacts and biases from measurement
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::84f055667aacece080262c17a0a2072c
https://doi.org/10.37917/ijeee.18.2.2
https://doi.org/10.37917/ijeee.18.2.2
Autor:
Seyedamir Shobeiri, Mojtaba Aajami
Publikováno v:
American Journal of Science & Engineering. 2:9-14
Deep learning models are in dire need of training data. This need can be addressed by encouraging data holders to contribute their data for training purpose. Data valuation is a mechanism that assigns a value reflecting a number to each data instance
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ce6a1d2525b86298bbef39aa94b4b553
https://doi.org/10.15864/ajse.2302
https://doi.org/10.15864/ajse.2302