A modified metaheuristic algorithm-integrated ELM model of cancer classification.

In a speedily degrading environment, cancer is acknowledged as the most menacing disease whose death rate is higher than others. For this reason, a number of researchers have analyzed the cancer-inducing genes and designed an effcient classification model to diagnose cancer effectively and quickly. In this study, random parameters of Extreme Learning Machine (ELM) were optimized through Self-Adaptive Multi-Population-based Elite strategy Jaya (SAMPEJ) approach. This strategy constructs a robust classifier called SAMPEJ-ELM model. This model was tested on datasets of breast, cervical, and lung cancers. To this end, a comparative analysis of the proposed model with ELM, Jaya optimized ELM (Jaya-ELM), SAMPEJ optimized Neural Network (SAMPEJNN), Teaching Learning Based Optimization (TLBO) hybridized ELM (TLBO-ELM), and SAMPEJ optimized Functional Link Artificial Neural Network (SAMPEJ-FLANN) models was conducted. Numerous performance metrices namely the accuracy, specificity, Gmean, sensitivity, and F-score with Receiver Operating Characteristic (ROC) were employed to evaluate the proposed approach. Moreover, this model was compared with 11 existing models. Of note, SAMPEJ-ELM approach had the highest degree of accuracy, sensitivity, and specificity in the datasets of breast (0.9895, 1, 0.9853), cervical (0.9822, 0.9948, 0.9828), and lung cancers (0.9787, 1, 1). The experimental outcomes revealed that SAMPEJ-ELM approach could classify the benign and malignant samples of cancer datasets significantly better than others. [ABSTRACT FROM AUTHOR]

Copyright of Scientia Iranica. Transaction D, Computer Science & Engineering & Electrical Engineering is the property of Scientia Iranica and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)