Classification of Benign and Malignant Breast Tumor Based on Machine Learning and Feature Selection Algorithms

This paper proposes a model that uses various machine learning classifiers and feature selection techniques to accurately predict breast cancer. The proposed model consists of three main stages: data pre-processing, feature selection, and finally different classifiers. During the pre-processing phase, missing values are addressed and the data is normalised. Subsequently, three different techniques are employed to select the most crucial features: mutual information, chi-squared, and Pearson correlation. The result of the feature selection stage is then fed into a number of classifiers, such as K-Nearest Neighbour (KNN), Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), and Logistic Regression (LR). The logistic regression classifier performs better in the experiments than the other methods, as evidenced by its accuracy of 97.08%, precision of 96.61%, sensitivity of 95%, and F-measure of 95.79%. Additionally, the suggested model is compared against cutting-edge algorithms, and the results demonstrated its superior performance.