Abstract:

Abstract Biomedical and health-care informatics research is increasingly using big data technology. At an unprecedented velocity and scale, large volumes of biological and clinical data have been created and gathered. DNA microarray classification has been widely used in biological and medical research to study gene expression patterns, identify disease biomarkers, classify cancer subtypes, predict treatment responses, and discover novel gene functions. Predictive analytics is becoming more popular for its applications in healthcare and has a lot of potential. While the performance concerns are still in a way and optimization approaches are used to address these concerns. In the proposed methodology we have adopted hybrid approach of optimization algorithm (Artificial Bee colony) for feature selection and deep learning (Convolutional Neural Network) method for classification. ABC method helps in obtaining best features which also improves the accuracy of classification.  The accuracy and other performance characteristics of the proposed algorithm CNN are examined. To demonstrate the usefulness of the proposed model, it is compared to various algorithms such as decision tree, random forest, and KNN based on performance metrics and the proposed approach achieves 98% accuracy which is remarkable as compared to other approaches.

Keywords: