Optimized Classification System for Structured Data

B. Suchitra and Dr.S. Duraisamy

Efficient classification of Clinical data is essential as they a have a greater impact on mankind. In this paper, a system is propounded to effectively select features, classify and optimise the classified data. Sample Clinical datasets are taken for study from clinical trials dataset. The structured data is pre-processed and features are extracted using Genetic Algorithm (GA). Classification is performed using Support Vector Machine (SVM). The classified data are optimised using Simulated Annealing (SA). The systems are analysed in terms of Maximum Accuracy, Average Accuracy, Specificity and Sensitivity. The proposed GA-SVM-SA offers 94.34%, 91.94%, 84.47%, 86.16% and 82.43% Accuracy for Congenital Adrenal Hyperplasia, Lead Poisoning, Cancer, Rheumatic Diseases and Heart Defects, Congenital datasets respectively.

Volume 11 | 04-Special Issue

Pages: 1421-1435