Breast Cancer Detection Using Multi Machine Learning Models

Abstract— Breast cancer prognosis remains an important area in health care. Breast cancer affects individuals of both sexes, but a higher prevalence is observed in the female population. To reduce this rate, machine learning techniques are being used, which have clear potential to improve the accuracy of cancer sensitivity and treatment. This paper presents a comparative study of the effectiveness of various machine learning algorithms with principal component analysis feature selection in the study to enhance model accuracy, and the same machine learning algorithm without principal PCA is used for breast cancer prediction. The application of PCA as a dimension reduction technique includes comparative analysis of common machine learning algorithms such as random forest (RF), support vector machine (SVM), logistic regression, and K-Nearest Neighbor models. Evaluation criteria including accuracy and sensitivity are used to evaluate the prediction performance of each method. In addition, it explores the effects of PCA on feature representation and model interpretation to provide insight into the trade-off between prediction accuracy and dimensionality reduction. The results of the PCA implementation generated by different ML algorithms which ranked SVM as the best classifier, obtained classification accuracies of 99.1%, 97.0%, and 99.3% on the three sets 75%-25%, 70%-30%, and 80 respectively sequence, although the implementation outcomes without PCA changed to... The classification accuracy was 97.9%, 97.6%, and 97.3% on the three sets 75%-25%, 70%-30%, and 80%- 20 respectively for the LR method, at Between methods of studying classification method, through a systematic comparison.