Feature selection and classification of microarray data are the most important challenges in machine learning. The motivation behind the Feature selection techniques is in selecting discriminate feature subsets which plays a vital role in the process of classifying cancer/tumour microarray expression data. In the present work, a novel feature selection approach is employed which combines F-Score and Relevant Information Gain (RIG) in miRNA data normalized by fuzzy Gaussian membership function. The F-score is employed to identify the discriminative features. The RIG is computed based on the class specific features of mean score values of the features. The experiments are conducted on seven miRNA datasets to demonstrate the performance of a proposed algorithm using the classifiers Support Vector Machine (SVM) and Artificial Neural Network (ANN). The experimental results show that the proposed approach gives a better classification accuracy compared to the state-of-the art feature selection algorithms. The proposed feature selection method gives 100% average classification accuracy with SVM and ANN for the Angulo_DI miRNA dataset and higher average classification accuracy for the other datasets compared to existing feature selection methods.