As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
Convolutional Neural Networks (CNNs) have been broadly used for image classification. However, with the exponential growth of digital images cause overloading problem which contains increasingly redundant, irrelevant features and noisy data which cause CNNs running gradually and in the meantime its classification accuracy decreasing. In this paper, we propose a 2D-reduction algorithm as data pre-processor for CNNs by using rough set theory with no information loss. 2D-reduction implies that the reduction of data in ways: feature reduction and noisy sample reduction. In the feature reduction stage, removing the set of irrelevant features by using rough set approach and for noisy sample reduction, we proposed novel Rough-KNN Noise reduction algorithm for reduction of noisy samples. The Rough set could adequately select the noisy boundary samples to eliminate on the basis of KNN rules, whose classes have been mislabeled. Experimental results demonstrate the proposed 2D-reduction approach can adequately reduce most of the irrelevant features & noisy samples, which causes the better generalization performance of CNNs.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.