Cloud Security: A Machine Learning Approach To Intrusion Detection

  • Unique Paper ID: 167431
  • Volume: 11
  • Issue: 3
  • PageNo: 1003-1012
  • Abstract:
  • Cloud computing provides on-demand access to a broad range of network and computer resources, encompassing storage, data management services, computing power, applications, and more. Users can easily access and utilize these resources as needed. The project focuses on enhancing cloud security by implementing an intrusion detection model leveraging machine learning techniques. The primary aim is to monitor and analyze resources, services, and networks within the cloud environment to effectively detect and prevent cyber-attacks. The proposed intrusion detection model utilizes machine learning techniques, specifically emphasizing the use of the Random Forest (RF) algorithm. Random Forest is a powerful ensemble learning method that combines multiple decision trees to make more accurate predictions. Feature engineering is a critical aspect of the model development process. It involves selecting and optimizing relevant features from the dataset to feed into the machine learning model. Effective feature engineering contributes to the model's ability to discern patterns and identify potential attacks accurately. The model's implementation is aimed at improving cloud security by continuously monitoring cloud resources, services, and networks. By applying machine learning algorithms, the model identifies unusual activities or patterns associated with cyber-attacks, thereby enhancing the overall security posture of the cloud infrastructure. The model's performance is evaluated and validated using two datasets: Bot-IoT and NSL-KDD. These datasets are common benchmarks in the field of intrusion detection. The model demonstrates high accuracy in detecting intrusions compared to recent related works, indicating its effectiveness and reliability in identifying potential security threats. The project's includes a Voting Classifier combination of RF + ADaBoost and Stacking Classifier with RF + MLP with LightGBM got 99% and 100% of accuracy for Kdd-Cup data respectively for enhanced cloud detection performance.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 3
  • PageNo: 1003-1012

Cloud Security: A Machine Learning Approach To Intrusion Detection

Related Articles