A hybrid approach for improving data classification based on PCA and enhanced ELM Online publication date: Thu, 03-Nov-2022
by Doaa L. El-Bably; Khaled M. Fouad
International Journal of Advanced Intelligence Paradigms (IJAIP), Vol. 23, No. 3/4, 2022
Abstract: The efficient and effective process of extracting the useful information from high-dimensional data is worth studying. High-dimensional data is big and complex and that it becomes difficult process and classify. Dimensionality reduction (DR) is important and the key method to address these problems. This paper presents a hybrid approach for data classification constituted from the combination of principal component analysis (PCA) and enhanced extreme learning machine (EELM). The proposed approach has two basic components. Firstly, PCA as a linear data reduction, is implemented to reduce the number of dimensions by removing irrelevant attributes to speed up the classification method and to minimise the complexity of computation. Secondly, EELM is performed by modifying the activation function of single hidden layer feedforward neural network (SLFN) for the perfect distribution of categories. The proposed approach depends on a static determination of the reduced number of principal components. The proposed approach is applied on several datasets and its effectiveness is supported by the different experiments performed. For more reliability, the proposed approach is compared with two of the previous works, which used PCA and ELM in data analysis.
Online publication date: Thu, 03-Nov-2022
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Advanced Intelligence Paradigms (IJAIP):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email email@example.com