Cyber Attacks Detection Using AI Algorithms

This research is analysing multiple artificial intelligence algorithms to detect cyber attacks

Check Report Status Status: draft, Type: Project

Victor Adankai, su21-reu-365, Edit

Keywords: AI, ML, DL, Cybersecurity, Cyber Attacks.

1. Introduction

  • Find literature about AI and Cyber Attacks on IoT Devices Dectection.
  • Analyze the literature and explain how AI for Cyber Attacks on IOT Devices Detection are beneficial.

Types of Cyber Attacks

  • Denial of service (DoS) Attack:
  • Remote to Local Attack:
  • Probing:
  • User to Root Attack:
  • Adversarial Attacks:
  • Poisoning Attack:
  • Evasion Attack:
  • Integrity Attack:
  • Malware Attack:
  • Phising Attack:
  • Zero Day Attack:
  • Sinkhole Attack:
  • Causative Attack:

Examples of AI Algorithms for Cyber Attacks Detection

  • Convolutional Neural Network (CNN)
  • Autoencoder (AE)
  • Deep Belief Network (DBN)
  • Recurrent Neural Network (RNN)
  • Generative Adversal Network (GAN)
  • Deep Reinforcement Learning (DIL)

2. Datasets

  • Finding data sets in IoT Devices Cyber Attacks.
  • Can any of the data sets be used in AI?
  • What are the challenges with IoT Devices Cyber Attacks data set? Privacy, HIPPA, Size, Avalibility
  • Datasets can be huge and GitHub has limited space. Only very small datasets should be stored in GitHub. However, if the data is publicly available you program must contain a download function instead that you customize. Write it using pythons request. You will get point deductions if you check-in data sets that are large and do not use the download function.

3. Using Images

  • Place a cool image into projects images in my directory
  • Correct the following link, replace the fa number with my su number and thne chart of png.
  • If the image has been copied, you must use a reference such as shown in the Figure 1 caption.

Figure 1

Figure 1: Images can be included in the report, but if they are copied you must cite them 1.

4. Benchmark

Your project must include a benchmark. The easiest is to use cloudmesh-common [^2]

5. Conclusion

A convincing but not fake conclusion should summarize what the conclusion of the project is.

6. Acknowledgments

  • Gregor von Laszewski
  • Yohn Jairo Bautista
  • Carlos Theran

7. References

  1. Gregor von Laszewski, Cloudmesh StopWatch and Benchmark from the Cloudmesh Common Library, [GitHub] ↩︎