In our study, we proposed a novel Neuro-fuzzy classification technique. The inputs to the Neuro-fuzzy classification system were fuzzified by applying the Gaussian curve membership function. The proposed method considered a fuzzification matrix in which the input features were associated with a degree of membership to different classes. Based on the value of degree of membership a feature would be attributed to a specific category or class. We applied our method to five benchmark datasets from the UCI machine language repository for classification. Our objective was to analyze the proposed method and, therefore compare its performance with Multilayer Perceptron Backpropagation Network (MLPBPN) algorithm in terms of different performance measures like Accuracy, Root-mean-square error, Kappa statistic, True Positive Rate, False Positive Rate, Precision, Recall, and F-Measure. In every aspect the proposed method performed better than MLPBPN.
ATA mining has attracted many researchers and analysts in the information industry and in research organizations as a whole in the last decades, due to the availability of large amounts of data and the impending need for changing such data into meaningful information and knowledge. The useful information and knowledge gained can be used for applications ranging from market survey, customer retention, and production control to evolutionary analysis and science exploration , .
Classification as an important data mining technique involves extracting interesting patterns representing knowledge from large real-world databases. Such analysis can provide a deep insight into the better understanding of different large-scale databases. The study related to effective knowledge development is also very popular in any research as because the decision-making process mainly depends upon the effectiveness of the classification method being utilized.
Basically data classification ,  is the method of discovering a model or classifier that describes and differentiates data classes so that the model could predict the class of entities with unknown class label value. It is a two-step procedure, in the first step; a classifier is constructed denoting a predefined set of concepts or data classes. This is the training phase, where a classification algorithm constructs the classifier by learning from a training data set and their associated class label attributes. In the next step the model is used for classification. In order to estimate the performance of the classifier a test set independent of the training tuples is used. Several pre-processing steps like data cleaning, data selection and data transformation are also applied to the data set before the actual classification procedure takes place.
Neural Network (NN) , ,  is a popular data modeling tool that can perform intelligent tasks similar to the human brain. NN is well-known for high precision and high learning ability even when a very little...