Modelo de identificación y clasificación de ataques de intrusión basado el conjunto de datos de “detección de intrusión en IT” del repositorio KAGGLE mediante la implementación de un algoritmo evolutivo

Abstract

This study develops a model for the identification and classification of intrusion attacks in IoT networks using the “Intrusion Detection in IT” dataset from the KAGGLE repository, by implementing an evolutionary algorithm. The main objective was to improve the accuracy of the intrusion detection model by optimizing a deep neural network using a genetic algorithm. The research was performed in a processing environment based on online code execution platforms, which provided advanced capabilities for handling large data volumes and complex models. The focus was on evaluating how data preprocessing, dimensionality reduction, and the application of an evolutionary algorithm influence the effectiveness of the deep neural network to classify intrusion attacks.