Few-shot classification using tensor completion
Автор: Chemometrics & Machine Learning in Copenhagen
Загружено: 2024-02-27
Просмотров: 274
Описание: Machine learning is increasingly applied to tackle classification challenges in various real-world contexts. Recently, there has also been a growing interest in tensor-based modeling techniques within the machine learning community. In this work, we combine classification and tensor decomposition methods to reformulate the classification problem as a tensor completion task. This tensor-based learning approach can enhance classification performance, especially when dealing with a constrained number of training samples, outperforming state-of-the-art methods.
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке: