A Hybrid Model Particle Swarm Optimization Based Mammogram Classification Using Kernel Support Vector Machine

Identifying affected cancer cells in women's breasts is mammogram, which is the major issue in the field of medicine all over the world. In order to raise the endurance of patients, it is most essential to identify the issue as early as possible. It also helps them to inflate the different options for treatment. With the new dramatic development in computation, machine learning made a revolution with dataset includes huge volume of breast images which could assist in recognizing malignant tumor with better diagnostics. Digital mammography images are taken, in that the x-ray images are read and stored in computer such that data can be easily enhanced and classified for further action. A novel approach is proposed in this paper to diagnose cancer affected cells with a good accuracy rate. Classification of mammogram with hybrid model includes feature extraction, various kinds of features are extorted from the intensity mammogram. A Particle Swarm Optimization optimizer is used in this paper which selects the features, and kernel-based Support Vector Machine classifier classifies the cancer lump from the taken mammogram metaphors. The exactness of a specific model can be assessed by the level of right forecasts made by the model.