Preprocessing Information from a Data Network for the Detection of User Behavior Patterns
This study focuses on the preprocessing of information for the selection of the most significant characteristics of a network traffic database, recovered from an Ecuadorian institution, using a method of classifying optimal entities and attributes, with the In order to achieve a complete understanding of its real composition to be able to generate patterns and identification of trends of behavior in the network, both of patterns that deviate from normal traffic behavior (intrusive), as well as normal, to detect with high precision possible attacks. Network management tools were used as a multifunctional security server software, as well as pre-processing of data tools for the selection of attributes, as well as the elimination of noise from the instances of the database, It allowed to identify which ins- tances and attributes are correct and contribute with effective information in the study. Among them we have: Greedy Stepwise Algorithm (Algoritmo Voráz), K-Means Algorithm, Discrete Chi-square Attributes and the use of computational models as Evolutionary Neural Networks and Gene Algorithms.
KeywordsIntrusion detection Server GreedyStepwise Algorithm K-Means Evolutionary neural networks Genetic algorithms
- 2.Reddy, R.R., Ramadevi, Y., Sunitha, K.V.N.: Effective discriminant function for intrusion detection using SVM. In: 2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI), pp. 1148–1153 (2016)Google Scholar
- 4.Farnaaz, N., Jabbar, M.: Random Forest Modeling for Network Intrusion Detection System (2016)Google Scholar
- 7.Yao, X.: Evolving artificial neural networks (1999)Google Scholar
- 9.García Vallejo, C.A.: Selección de Instancias y Atributos en Conjuntos de Datos mediante Algoritmos sobre Grafos (2012)Google Scholar
- 10.Morring, B.D., Martinez, T.R.: A nearest neighbor data reduction algorithm (2004)Google Scholar