Cyber AI training practically examines the following two issues.
- Can you hack Artificial Intelligence? (Cyber AI)
- Can you hack Artificial Intelligence? (AI Security)
With the increase in the use of technology and software products in our living spaces, the need for cyber security is increasing exponentially. The increase in the number and variety of these products also causes the attack surface in cyber security terminology to expand. For this reason, many technology initiatives, institutions and organizations aim to solve various security problems by developing cyber security software. However, in many cyber security scenarios where 'self-learning' systems are required, traditional software is insufficient. For this reason, the use of artificial intelligence and cyber security together is becoming more and more common in the world.
As with all new concepts, the concept that deals with solving cybersecurity problems with artificial intelligence has many names. We call it 'Cyber AI' in general (it stands for Cyber Security AI). One of the topics we will cover in this course is Cyber AI and the other is AI Security.
So, what is AI Security?
All artificial intelligence systems can be hacked!
Just as there are domain-specific security problems in the web, mobile, database or any other field, artificial intelligence has its own security problems. In artificial intelligence, there are multiple attack techniques that will allow hacking of the relevant artificial intelligence algorithm, regardless of which title we examine, whether text, sound, frequency or image. Of course, their defense methods... In this training, we will examine both 'Cyber Security AI', which includes artificial intelligence in cyber security problems that have existed for years, and 'AI Security', which deals with the security vulnerabilities of artificial intelligence itself.
- Basic knowledge of Python programming language.
- To have basic level knowledge in the field of cyber security.
*Detailed information about the basics of artificial intelligence will be given in this training. Therefore, knowledge of artificial intelligence is not a prerequisite.
Cyber AI training is completely hands-on and targets cybersecurity professionals.
If you have a good command of the basics of cyber security and can do basic programming with Python, you have the basic requirements for participation in this training.
Technologies to be used
- Python
- CUDA, Flask,
- PyTorch, TensorFlow, Keras, OpenCV
- AI Security Tools/Tooling
- AI-Based Cyber Security Tools/Tooling
- NumPy, scikit-learn, Pandas, ONNX, Matplotlib and dozens of different libraries/tools...
- Visual Studio Code, Jupyter Notebook, Google Colab
Outline
01 - Artificial Intelligence Application Development Fundamentals
Artificial Intelligence Application Development Overview
Development Tools
NumPy, TensorFlow and PyTorch
Importance of NumPy : Computational Intelligence, TensorFlow and Relationship with PyTorch
Numerical Computing with NumPy
Data Manipulation with Pandas
Programming with TensorFlow
Programming with PyTorch
Machine Learning vs. Deep Learning
Machine Learning
Machine Learning Fundamentals
Project : Machine Learning Application
Computer Vision
Computer Vision Fundamentals
Project : Machine Learning Application
Deep Learning
Deep Learning Fundamentals
Project : Deep Learning Application
02 - Cyber Security with Artificial Intelligence (Cyber AI)
Cyber Security Overview with Artificial Intelligence
Use Scenarios of Artificial Intelligence in Cyber Security
The Limits of Artificial Intelligence in Cyber Security
Advantages and Disadvantages of Artificial Intelligence in Cyber Security
Malicious URL Detection with Artificial Intelligence
Theoretical Explanation
Project : Malicious URL Detection
Network Anomaly Detection Application
Theoretical Explanation
Project : Network Anomaly Detection
Log Analysis with Artificial Intelligence
Theoretical Explanation
Project : Log Analysis
Phishing URL/Website Detection with Artificial Intelligence
Theoretical Explanation
Project : Phishing URL/Website Detection
XSS Detection with Artificial Intelligence
Theoretical Explanation
Project : XSS Detection
Credit Card Fraud Detection with Artificial Intelligence
Theoretical Explanation
Project : Credit Card Fraud Detection
Static Code Analysis with Artificial Intelligence
Theoretical Explanation
Project : Static Code Analysis
SQL Injection Detection with Artificial Intelligence
Theoretical Explanation
Project : SQL Injection Detection
Code Similarity with Artificial Intelligence
Theoretical Explanation
Project : Code Similarity
Steganography with Artificial Intelligence
Theoretical Explanation
Project : Steganography
Captcha Breaker with Artificial Intelligence
Theoretical Explanation
Project : Captcha Breaker
XSS Payload Generation with Artificial Intelligence (XSS Attacker)
Theoretical Explanation
Project : XSS Payload Generation
DDoS Attack Detection with Artificial Intelligence
Theoretical Explanation
Project : DDoS Attack Detection
Finding Credential (Password) in a File with Artificial Intelligence
Theoretical Explanation
Project : Artificial Intelligence Finding Passwords in Files
Malware Detection with Artificial Intelligence
Theoretical Explanation
Project : Malware Detection
03 - Cyber Security for Artificial Intelligence (AI Security)
AI Security Overview
What is an Adversarial Attack?
Artificial Intelligence Hacking Scenarios with Adversarial Attack
White Box vs. Black Box
Overview of AI Security Vulnerabilities
Perturbation Attack
Poisoning Attack
Model Inversion
Membership Inference
Model Stealing
Reprogramming Machine Learning System
Adversarial Example in Physical Domain
Malicious Machine Learning Provider Recovering Training Data
Attacking the Machine Learning Supply Chain
Backdoor Machine Learning
Exploit Software Dependencies
Reward Hacking
Side Effects
Distribution Shifts
Natural Adverse Examples
Common Corruption
Incomplete Testing
Artificial Intelligence Security Applications
04 - Data Privacy, Federated Learning and Encrypted Machine Learning
Data Privacy
- Data Privacy Basics
- Why Should We Care About Data Privacy?
- Ways to Increase Privacy
- Which Data Should Be Confidential?
- TensorFlow Privacy
Federated Learning
- What is Federated Learning and Why Is It Used?
- Federated Learning Architecture
Encrypted Machine Learning
- What is Encrypted Machine Learning and Why Is It Used?
- Enrypted Machine Learning Architecture
- Encrypted Model Training
- Encrypted Prediction
Join our public courses in our Finland facilities. Private class trainings will be organized at the location of your preference, according to your schedule.