Author Login IJIRT Partner Network (IPN) Login

Deep Learning based Network Anomaly Detection

  • Unique Paper ID: 169206
  • Volume: 11
  • Issue: 6
  • PageNo: 798-807
  • Abstract:
  • Because it offers a practical way to stop and combat network intrusions, network anomaly detection is essential. Many deep learning techniques based on Autoencoders (AUTOENCODERs) have been created for anomaly identification as a result of advances in Artificial Intelligence (AI). The efficacy of the current AUTOENCODER models varies, nevertheless, and they lack a thorough method for evaluating important performance indicators and maximizing detection accuracy. A unique 5 layer Autoencoder model designed for identification of network anomalies is presented in this paper. Our strategy is based on a careful analysis of performance parameters that are crucial to AUTOENCODER-based identification. Our platform uses a novel data processing strategy that alters and eliminates outliers that significantly impact the feature set in order to overcome potential biases from unbalanced data. Our approach consistently differentiates between normal and abnormal using an improved reconstruction error function and abnormal network traffic. Our model delivers improved detection accuracy and F1-score when combined with an efficient feature learning and dimensionality reduction architecture. Our suggested model performs better than alternative approaches when tested on dataset, with an F1-score of 92.26% and a detection accuracy of 90.61%.
Download article email to a friend

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 6
  • PageNo: 798-807

Deep Learning based Network Anomaly Detection

Related Articles

18 Apr, 2024

Neural Network Wardrobe Consultant based on ResNet50

Read More
Impact Factor
8.01 (Year 2024)
UGC Approved
Journal no 47859

Join Our IPN

IJIRT Partner Network

Submit your research paper and those of your network (friends, colleagues, or peers) through your IPN account, and receive 800 INR for each paper that gets published.

Join Now

Latest Publication

Archive

View More

Recent Conferences

NCSEM 2024

National Conference on Sustainable Engineering and Management - 2024 Last Date: 15th March 2024

Submit inquiry