This paper proposes an experimental methodology
based on the comparison of several deep learning models for the
detection of distributed denial of service (DDoS) attacks in
enterprise network environments. The CIC-DDoS2019 dataset,
recognized for the richness and realism of its attack scenarios,
served as a basis for the preparation, training and evaluation of
the models. Five deep learning architectures were developed and
tested according to a binary classification strategy, distinguishing
normal traffic from malicious traffic. Performance was evaluated
using standard metrics: Accuracy, Precision, Recall, F1-score and
Confusion Matrix. The results obtained demonstrate the overall

effectiveness of the studied models, in particular the CNN-
BiLSTM, which presents the best balance between accuracy,

stability and generalization capacity when faced with new data.

Cybersecurity, Deep Learning, Binary Classification, Intrusion Detection System (IDS), DDoS, CIC- DDoS2019 dataset