Detection of Malicious Bots Using a Proactive Supervised Classification Approach

Malicious bots are one of the most commonly used tools by cybercriminals today to carry out security breaches. These malicious programs can simulate human activity, which is the reason why they affect t a large number of websites. Different techniques have been developed to ensure the detection of malicious bots, highlighting the application of Machine Learning algorithms and the meta-learning branch, for improving their performance. The present work proposes the application of the Proactive Forest algorithm as the main element of a module that allows the detection of malicious bots, based on Machine Learning. An experimental study was carried out to measure its performance, based on a comparison with the Random Forest algorithm. The results showed that 99.93% of instances were correctly classified and 99.96% were correctly classified as malicious bots out of the total, as the best results achieved.