Hybrid Artificial Bee Colony Algorithm and Semi Supervised Learning Prediction Model for the Risk of Cardiovascular Disease in Type-2 Diabetic Patients
P. Radha1, B. Srinivasan2
1P. Radha, Ph.d Scholar, Department of Computer Science, Karpagam University, Coimbatore, Asst. Prof., Vellalar College for Women, Erode, India.
2Dr. B. Srinivasan, Department of Computer Science, Gobi Arts and Science College, Gobichettipalayam, India
Manuscript received on 30 January 2015 | Revised Manuscript received on 12 February 2015 | Manuscript Published on 28 February 2015 | PP: 17-24 | Volume-4 Issue-9, February 2015 | Retrieval Number: I1971024915/15©BEIESP
Open Access | Editorial and Publishing 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: Cardiovascular disease (CVD) factor is one of the important causes of mortality among diabetes patients. Statistics shows that more than 22% of people with type 2 diabetes mellitus suffer from CVD and which in turn leads to cardiovascular disease. But still some of the works doesn’t mainly focus on the semisupervised learning methods with feature selection methods to enhance the prediction accuracy of the classification methods. The aim of this research was to identify significant CVD factors influencing type 2 diabetes controls to improve prediction accuracy. In proposed methods the preprocessing and dimensionality reduction of the patients records is done by using Kullback Leiber Divergence(KLD) –Principal component analysis (PCA) ,then attribute values measurement is completed by using kernel density estimation (KDE) which measures the attributes values based on probability mass function with Gaussian kernel function, feature selection is performed by using artificial bee colony with differential evolution (ABC-DE). Hybrid prediction model Improved Fuzzy C Means (IFCM) clustering algorithm aimed at validating chosen class label of given data and subsequently applying semisupervised Modified Self-Organizing Feature Map Neural Network (MSOFMNN) classification algorithm to the result set. The proposed method examines the behavioral factors that contribute to CVD risk factors among those with type 2 diabetes (T2D) with higher prediction accuracy, less error rate.
Keywords: Artificial bee colony (ABC), Classification, Hybrid Prediction Model, Kernel density estimation (KDE), Modified Self-Organizing Feature Map Neural Network (MSOFMNN).
Scope of the Article: Neural Network