Evaluation of Biometric Classification and Authentication Using Machine Learning Techniques

This research article proposed the performance measurement of biometric image with computational methodology. This research adopts the following procedures: pre-processing, Feature extraction and Classification. Designing and building the algorithm and simulation programs have been done in a MATLAB environment. My SQL has been utilized for the maintenance of overall dataset. By employing the classification technique comparative analysis was carried out by inspecting three data mining techniques which were: Random Tree, The multilayer Perceptron neural network (MPNN), and the C4.5 decision tree (DT) algorithms. As per the final conclusion, the Random Forest Classifier algorithm exhibits greater performance in contrast to the other techniques. It has been found that the 93.5 % accuracy exhibited by Random Forest Classifier is much greater and enhanced.