CNN-BD: An Approach for Disease Classification and Visualization

Abstract

Data visualization is one of the complex parts of the discovery process in the current phase of big data. Finding the hidden level data of big data is that the principal goal of the classifier. The size of the data, number of classes, and the feature space had an effect on the performance of the classifiers. The new analysis of algorithms is needed for improving the accuracy, efficiency, and reliability of the classifiers. This paper proposes a Deep Learning based Convolution Neural Network classifier to classify and visualize the disease data. PCA and PSO methods are used for multivariate data analysis to handle massive data and feature selection. To demonstrate the proposed learning algorithm, real-world datasets are used. The comparative study shows that deep learning classifier performs better than other classifiers and scientifically higher.