Detection of Liver disorder using Quadratic Support Vector Machine in comparison with RBF SVM to measure the accuracy, Precision, sensitivity and specificity

Abstract

The purpose of this study is to compare the Quadratic SVM classifier’s performance with the RBF SVM method in identifying liver disorders. Techniques and Resources: Three datasets on liver illness that are available in Kaggle contain a total of 31035 samples. These samples are split into two groups: the training dataset (n = 23276; 75% of the total) and the test dataset (n = 7759; 25% of the total). Values for accuracy, precision, specificity, and sensitivity are calculated to estimate the SVM algorithm’s performance. Results: The accuracy, precision, sensitivity, and specificity of the quadratic SVM algorithm were 73.60 percent, 99.89 percent, 73.01 percent, and 96.87 percent, respectively, as opposed to the RBF SVM algorithm’s 73.32 percent, 97.97 percent, 77.27 percent, and 70.08 percent. In this research, it was discovered that the Quadratic SVM algorithm outperformed the RBF SVM algorithm in liver.