A Comparative Study of Seven Machine Learning Algorithms for Breast Cancer Detection and Diagnosis
<p>Qinyi Ruan</p> · Academic Journal of Medicine & Health Sciences · 2023
This paper presents a comparative analysis of seven distinct machine learning (ML) algorithms, namely Linear Discriminant Analysis, Logistic Regression, K-Nearest Neighbor, Decision Tree Classifier, Random Forest Classifier, Voting Classifier, and Support Vector Machine, in predicting the diagnosis of breast cancer. The study utilized 30 histological tumor features obtained from digital imaging of fine needle aspirates of breast tumor cell masses contained in the dataset, achieving an accuracy of approximately 95% through the application of the aforementioned algorithms. Results show that the LDA and RFC algorithms outperformed the others in terms of accuracy in diagnosing breast cancer. Furthermore, the study suggests that the stability of diagnostic outcomes is better achieved with large-scale data. Finally, the accuracy of the LR algorithm was observed to be less than 85% after conducting Principal Component Analysis (PCA), which was lower than the accuracy achieved without dimensionality reduction.