Predicting Breast Cancer Using Changing Parameters of Machine Learning Model

Abstract

AbstractBreast cancer is the most common type of fatal ailment seen in females in the world. There are many types of cancer, one of which is breast cancer. Several types of breast cancer are found in women that affecting their lives across the world. Several types include “lobular carcinoma in situ (LCIS), ductal carcinoma in situ (DCIS), invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC).” How many people are dying due to cancer today, and one of the main reasons is not known in time. Generally, breast cancer may be a malignant neoplasm that begins within the cell of the breast and eventually spreads to the encompassing tissue. Due to breast cancer, a lot of death is happening. The death rate can be reduced by using machine learning techniques. Mammography is a good and effective modality that is used in the detection of breast cancer in today’s time. In this paper, we used different machine learning algorithms like Naïve Bayes, k-nearest neighbors, logistic regression, support vector machine, decision tree, and convolution neural network. After changing the unique hyperparameter of each model, find the better accuracy within the model and also do the comparison between models. The performance of convolution neural network is found maximum accuracy with minimum loss. The accuracy achieved by convolution neural network is 99.05%.KeywordsBreast cancerSupport vector machine (SVM)k-nearest neighbor (KNN)Naive BayesLogistic regressionDecision treeConvolution neural network