Empirical Processing of Breast Cancer Prediction Strategies using DEFS Algorithm
R. Preetha1, S. Vinila Jinny2
1R. Preetha*, Department of CSE, Noorul Islam Centre for Higher Education, Thuckalay, Kumaracoil-629180, Tamil Nadu, India.
2Dr. S. Vinila Jinny*, Department of CSE, Noorul Islam Centre for Higher Education, Thuckalay, Kumaracoil-629180, Tamil Nadu, India. 

Manuscript received on January 01, 2020. | Revised Manuscript received on January 20, 2020. | Manuscript published on January 30, 2020. | PP: 3081-3087 | Volume-8 Issue-5, January 2020. | Retrieval Number: E6288018520/2020©BEIESP | DOI: 10.35940/ijrte.E6288.018520

Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Now-a-days an important threat to women over global manner is Breast-Cancer, which is the major disease cause drastic affection to female especially. Identification of Breast Cancer over earlier stages is must to save one’s life and the significant affection range of Breast-Cancer is drastically improved day by day due to the improper food-habits, pollution-level and improper-life style as well as genetic-issues also. The main cause of this disease is the arising of breast-ample over the ‘breast-area, which develops the cancer to women in several cases. If the detection or prediction of such masses over earlier stage will helps to women to get more survival ratio as well as this leads a proportion to researchers to make an systematic process to detect such diseases on initial stages by using intelligent prediction methodologies with high accuracy rates. In this paper, the proposed system handles several stages of processing to make sure the prediction accuracy, such steps are as follows: Data acquisition, Feature vector formation by normalization, Feature Selection by using Differential Evolution based selection methodology, Classification using Subspace Ensemble Learning and different Performance Measures. By using these strategies the entire work assures the proposed system is perfect to predict or identify the breast cancer benign/malignant stages more accurately compare to the classical Margin-Based Feature-Selection process. Compared to the classical biopsy methodology, a systematic diagnosis attains more impact due to its prediction accuracy. This proposed system is powered by a powerful approach called Differential-Evolution Feature’-Selection (“DEFS”) with the association of Subspace Ensemble Learning Classification principle, which provides highest accuracy and prediction rates compare to the classical methodologies. This proposed paper assures effective and robust mining strategies in Breast Cancer identification/prediction as well as efficient decision-making norms. The proposed outcome proves the good accuracy and resulting levels by means of Precision-Recall, Sensitivity and Specificity, True Positive/True Negative, False Positive/False Negative, Accuracy and Time Consumption.
Keywords: Breast Cancer Prediction, DEFS, Differential Evolution Feature Selection, Subspace Ensemble Learning Classification.
Scope of the Article: Classification.