Optimized Feature Extraction and Hybrid Classification Model for Heart Disease and Breast Cancer Prediction
Sireesha Moturi1, S. N. Tirumala Rao2, Srikanth Vemuru3
1Sireesha Moturi, Department of Computer Science and Engineering, KLEF, Vijayawada, India.
2Dr.S.N.Tirumala Rao, of Computer Science and Engineering, Narasaraopeta Engineering College, Narasaraopet, India.
3Dr. Srikanth Vemuru, Department of Computer Science and Engineering, KLEF, Vijayawada, India
Manuscript received on 23 March 2019 | Revised Manuscript received on 30 March 2019 | Manuscript published on 30 March 2019 | PP: 1754-1772 | Volume-7 Issue-6, March 2019 | Retrieval Number: F2343037619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© 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: Nowadays, diseases like heart disease and breast cancer are rising day by day due to the life style, hereditary and so on. Particularly, heart disease has become more common these days, i.e. life of people is at risk. Each and every individual has various values for cholesterol, Blood pressure, pulse rate and so on. However, the prediction of heart disease with data mining classification is not up to the mark. Hence, this paper intends to propose a new disease prediction model with advanced and modified classification technique. The proposed prediction model includes three phases: Coalesce rule generation, Optimized feature extraction and hybrid classification. Initially, the given big data is preprocessed by transforming the data to some other form, from which the rules are generated. The optimal features are selected by a new introduced algorithm namely, New levy Update based Dragonfly Algorithm (NL-DA). Finally, the selected optimal features are subjected to the new hybrid classifier, hybridization of Support vector Machine (SVM) and Deep belief Network (DBN), so that the accurate disease prediction is worked out. The proposed NL-DA model is compared to other conventional methods in terms of Accuracy, Specificity, Sensitivity, Precision, F1Score, Negative Predictive Value (NPV) and Matthews Correlation Coefficient (MCC), False negative rate (FNR), False positive rate (FPR) and False Discovery Rate (FDR), and proven the betterments of proposed work.
Keywords: Disease Prediction; Data Mining; Feature Extraction; Optimization; Classification.
Scope of the Article: Classification