DAY 4 – NEXT-GEN Cybersecurity with AI

Machine Learning (ML) and Deep Learning (DL) both rely on training data to detect threats in cybersecurity systems

• True

• False

Adversarial attacks strengthen AI models by improving their detection accuracy

• True

• False

Which of the following is a common dataset used for training Intrusion Detection Systems (IDS)?

• ImageNet

• CICIDS2017

• MNIST

• COCO

What is one major disadvantage of Deep Learning (DL) models in cybersecurity?

• They require less data than ML models

• They are easily interpretable

• They act as rule-based systems

• They are less explainable and require large datasets

Which of the following describes an “Evasion Attack”?

• Inserting malicious samples into training data

• Slightly modifying malware to bypass detection

• Extracting a model’s logic via repeated queries

• Encrypting network traffic with TLS