Hybrid Swarm Intelligence Algorithms with Ensemble Machine Learning for Medical Diagnosis

Disease Diagnosis still an open problem in current research. The main characteristic of diseases diagnostic model is that it helps physicians to make quick decisions and minimize errors in diagnosis. Current existing techniques are not consistent with all diseases datasets. While they achieve a good accuracy on specific dataset, their performance drops on other diseases datasets. Therefore, this paper proposed a hybrid Dynamic ant colony system three update levels, with wavelets transform, and singular value decomposition integrating support vector machine. The proposed method will be evaluated using five benchmark medical datasets of various diseases from the UCI repository. The expected outcome of the proposed method seeks to minimize subset of features to attain a satisfactory disease diagnosis on a wide range of diseases with the highest accuracy, sensitivity, and specificity.