Search

Breast cancer is a malignant tumor that originates in the cells of the breast. It is the second leading cause of women’s death, after lung cancer. Moreover, the availability of medical data facilitates the development of related Artificial Intelligence Systems (AIS). The diagnosis (or classification) of breast cancer is a delicate task, which requires efficient and robust classifiers. However, classical classification methods (in which a single basic classifier ( estimator )) are generally confronted with the “bias-variance” dilemma. This, very often, affects seriously their efficiency and robustness. In this article, to mitigate this problem, we propose a new learning model called Triple-Stacking. This technique is composed of three (3) methods of statistical learning (Logistic Regression, Voting and Stacking) and a meta-learner (Decision Stump). The proposed model outperformed the existing ones on two different databases: Breast Cancer Wisconsin Original Data Set and Breast Cancer Wisconsin Diagnostic Data Set, with accuracies of 99.57% and 99.64%, respectively.

Breast cancer is a malignant tumor that originates in the cells of the breast. It is the second leading cause of women’s death, after lung cancer. Moreover, the availability of medical data facilitates the development of related Artificial Intelligence Systems (AIS). The diagnosis (or classification) of breast cancer is a delicate task, which requires efficient and robust classifiers. However, classical classification methods (in which a single basic classifier ( estimator )) are generally confronted with the “bias-variance” dilemma. This, very often, affects seriously their efficiency and robustness. In this article, to mitigate this problem, we propose a new learning model called Triple-Stacking. This technique is composed of three (3) methods of statistical learning (Logistic Regression, Voting and Stacking) and a meta-learner (Decision Stump). The proposed model outperformed the existing ones on two different databases: Breast Cancer Wisconsin Original Data Set and Breast Cancer Wisconsin Diagnostic Data Set, with accuracies of 99.57% and 99.64%, respectively.

Breast cancer is a malignant tumor that originates in the cells of the breast. It is the second leading cause of women’s death, after lung cancer. Moreover, the availability of medical data facilitates the development of related Artificial Intelligence Systems (AIS). The diagnosis (or classification) of breast cancer is a delicate task, which requires efficient and robust classifiers. However, classical classification methods (in which a single basic classifier ( estimator )) are generally confronted with the “bias-variance” dilemma. This, very often, affects seriously their efficiency and robustness. In this article, to mitigate this problem, we propose a new learning model called Triple-Stacking. This technique is composed of three (3) methods of statistical learning (Logistic Regression, Voting and Stacking) and a meta-learner (Decision Stump). The proposed model outperformed the existing ones on two different databases: Breast Cancer Wisconsin Original Data Set and Breast Cancer Wisconsin Diagnostic Data Set, with accuracies of 99.57% and 99.64%, respectively.