The Support vector machines are learning models that are supervised in nature. The SVM models are models with related learning algorithms. These learning models perceive examples, patterns, and investigate data given. The SVM models are generally used in classification and regression models. The SVM model works by training the examples or data to fall into either one of the two categories. Then the model is built using the SVM algorithm and the new entry is made to fall into either of the two categories. In simple terms, SVM model is portrayal of the samples as points in space. The points are mapped so that the illustrations of the different classes are isolated by a reasonable crevice that is as wide as could be expected under the circumstances. The new entries or cases are then represented in the same space as the previous trained examples. The new cases are then anticipated to have a place with a classification in light of which side of the crevice they fall on.We used Support Vector Machine as it is widely used for breast and prostate cancer diagnosis [29].An RBF kernel with optimized gamma value for SVM is chosen for the experiments [13].

 

[13] J. Yao, D. Ganti, X. Luo, G. Xiao, Y. Xie, S. Yan, and J. Huang, “Computer-assisted diagnosis of lung cancer using quantitative topology
features,” in Machine Learning in Medical Imaging, ser. Lecture Notes in Computer Science, L. Zhou, L. Wang, Q. Wang, and Y. Shi, Eds. Springer International Publishing, 2015, vol. 9352, pp. 288–295. [Online]. Available: http://dx.doi.org/10.1007/978-3-319-24888-2 35

[29] S. Doyle, S. Agner, A. Madabhushi, M. Feldman, and J. Tomaszewski, “Automated grading of breast cancer histopathology using spectral clustering with textural and architectural image features,” in IEEE International Symposium on Biomedical Imaging. IEEE, 2008, pp. 496–499.