Abstract- In this research, four linear SVM classifiers are used to rapidly diagnose breast cancer. SVM classifiers were used to analyze breast cancer using the WBCD informative index. There are various modified classifiers, such as Linear Programming SVM and Lagrangian SVM, that are compared to SVM in terms of classification performance. SVM surpassed all other algorithms for all exhibition lists, with an accuracy of 97.71 percent, whereas Lagrangian SVM had the lowest accuracy, at 95.61 percent. A distant second place goes to linear programming SVM (with an accuracy of 97.33 percent). In order to get the best results, it is essential that the classifier and kernel functions be determined. During the validation stage, Linear Programming SVM achieved an overall accuracy of 97.14 percent, outperforming Lagrangian SVM (95.43%), Proximal SVM (96 %), and SVM (94.86 %). The overall sensitivities of Linear Programming SVM accomplished 98.25%, which is better than Lagrangian SVM (96.52%), Proximal SVM (97.37%) and SVM (95.65%). The overall specificities of Linear Programming SVM accomplished (95.08 %), which is better than Lagrangian SVM (93.33%), Proximal SVM (93.44%) and SVM (93.33%). Estimation of AUC for LPSVM accomplished 99.38%, individually, which beat different classifiers. The outcomes firmly recommend that Linear Programming SVM can help in the analysis of cancer data.

Keywords- Classification, SVM, Proximal SVM, Lagrangian SVM, Linear Programming SVM