Clinical decision Support System (CDSS) aid in analyzing clinical data and using them to predict disease and supply necessary care. In this research work, a paradigm for predict the accuracy of majority classifier. Clinical data classification was performed using two datasets from the University of California, Irvine (UCI) machine learning repository, namely the Hypothyroid and Autistic Screening datasets, which were aided by feature selection using bio-inspired computational methods and Back Propagation Neural Network (BPNN). Bat Algorithm (BA) and Crow Search Algorithm (CSA) are used for feature selection in Wrapper approach method and Stochastic Gradient Descent with Back Propagation is used to evaluate as the fitness function