An investigation of genetic algorithm-based feature selection techniques applied to keystroke dynamics biometrics

Due to the continuous use of social networks, users can be vulnerable to online situations such as paedophilia treats. One of the ways to do the investigation of an alleged pedophile is to verify the legitimacy of the genre that it claims. One possible technique to adopt is keystroke dynamics analysis. However, this technique can extract many attributes, causing a negative impact on the accuracy of the classifier due to the presence of redundant and irrelevant attributes. Thus, this work using the wrapper approach in features selection using genetic algorithms and as KNN, SVM and Naive Bayes classifiers. Bringing as best result the SVM classifier with 90% accuracy, identifying what is most suitable for both bases.