Class specific dynamic feature selection technique — Towards human movement based biometrics application

Classification of the biometrics data for identity validation can be considered as a single-class problem. Each class can be represented by a unique set of features. However, current feature selection techniques consider the entire database and identify the feature-set that is suitable for representing all available classes. This may not be the best representation of the biometrics data of each individual because different people may have difference in the most suitable features to represent their biometric data. In this study, a class-specific dynamic feature selection method has been proposed and experimentally validated using dynamic signatures. This method is based on the variance within the feature set, where the features with smaller variance are selected and the ones with larger variance are rejected. A comparison was made with other feature selection methods, and the results show that there were differences in the features representing different classes. A significant improvement in the classification accuracy and specificity and sensitivity was also observed when using this feature selection technique.