Paper Info
Title
Oppositional firefly optimization based optimal feature selection in chronic kidney disease classification using deep neural network
Abstract
Nowadays, medical data classification plays an important role in healthcare informatics applications such as disease prediction, classification, etc. The recently developed machine learning and deep learning models are commonly employed for effective medical data classification, which can be applied for disease diagnosis. The existing techniques make few shortcomings in terms of computational complexity, higher-dimensional features, higher execution time, etc. To tackle these issues, this paper develops a new classification model with metaheuristic algorithm based optimal feature selection for Chronic Kidney Disease diagnosis. Initially, the data with missing values were evacuated in the pre-processing stage. Then, the best subset of features selected by a metaheuristic algorithm named Oppositional based FireFly Optimization algorithm, which helps in the prediction or classification of the disease more accurately. The incorporation of oppositional based learning concept helps to improve the convergence rate of FireFly algorithm. For classification, Deep Neural Network was proposed to diagnose the existence of CKD. The effectiveness of the proposed feature selection-based classifier was tested on the dataset as measures of sensitivity, specificity, and accuracy. The results concluded that the proposed algorithm achieved a high accuracy rate when compared to the algorithms of existing classifier models.
Year
DOI
Venue
2022
10.1007/s12652-021-03477-2
Journal of Ambient Intelligence and Humanized Computing
Keywords
DocType
Volume
Medical data classification, CKD dataset, OFFO algorithm, DNN classifier
Journal
13
Issue
ISSN
Citations 
4
1868-5137
0
PageRank 
References 
Authors
0.34
15
2
Name
Order
Citations
PageRank
Jerlin Rubini Lambert100.34
Eswaran Perumal201.01