Scribd
Upload
8 views6 pages

Nader Gamal WDP - Nader Gamal

The document contains multiple-choice questions and essays related to privacy and security in AI and blockchain technologies. Key topics include differential privacy, federated learning, generative AI models, and their implications for data protection and model training. It also discusses the role of privacy laws and techniques for enhancing privacy in AI systems.

Uploaded by

Saad Mohamed Saad
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
8 views6 pages

Nader Gamal WDP - Nader Gamal

The document contains multiple-choice questions and essays related to privacy and security in AI and blockchain technologies. Key topics include differential privacy, federated learning, generative AI models, and their implications for data protection and model training. It also discusses the role of privacy laws and techniques for enhancing privacy in AI systems.

Uploaded by

Saad Mohamed Saad
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Paper 1 MCQ (1-5)

1. What type of attack tries to determine if a specific data point was used in training a
model?
A) Data poisoning
B) Contextual attack
C) Membership inference
D) Model inversion

2. Which technique can help reduce risks from model memorization?


A) Selective forgetting
B) Overfitting
C) Data replication
D) Model expansion

3. What is a potential consequence of fine-tuning a model on sensitive data?


A) Model compression
B) Data redundancy
C) Improved accuracy
D) Information leakage

4. What can model inversion attacks reveal from generative AI models?


A) Training accuracy
B) Sensitive personal data
C) Hyperparameters
D) Model size

5. What approach helps organizations align generative AI use with legal obligations?
A) Open data sharing
B) Innovation-first development
C) Privacy by design
D) Model scaling
Paper 1 MCQ (6-10)
6. What does differential privacy primarily use to protect individual data?
A) Secure channels
B) Encryption keys
C) Noise injection
D) Data masking

7. Which technique enables decentralized model training without sharing raw data?
A) Secure MPC
B) Federated Learning
C) Data anonymization
D) Homomorphic Encryption

8. What is the main benefit of homomorphic encryption in AI?


A) Allows computation on encrypted data
B) Reduces training time
C) Increases data transfer speed
D) Improves data visualization

9. Which tool supports anonymization and masking of sensitive data in compliance


with GDPR?
A) TensorFlow
B) PyGrid
C) PyTorch
D) Microsoft Presidio

10. What is a major privacy concern when using synthetic data in sensitive fields like
healthcare and finance?
A) High storage costs
B) Risk of data re-identification
C) Overfitting to public datasets
D) Difficulty in generating realistic images
Paper 1 Essay (11-15)

11. How does differential privacy (DP) protect sensitive data?

It adds noise to data or results, making it hard to trace back to individuals.

12. How does federated learning (FL) improve privacy in AI training?

It trains models on local devices without sharing raw data.

13. What does “privacy by design” mean in AI development?

Building privacy features into systems from the start phases.

14. What role do privacy laws like GDPR play in AI development?

They require AI to protect personal data and ensure user privacy.

15. How can synthetic data help prevent data leakage in AI models?

It replaces real data, reducing the risk of exposing sensitive information.


Paper 2 MCQ (16-20)
16. What is a key advantage of GAI over DAI in blockchain?
A) Faster mining
B) Uses less power
C) Generates new data
D) Requires no training

17. Which GAI model denoises data to generate realistic samples?


A) GDM
B) GAN
C) VAE
D) LLM

18. What challenge does GAI help solve in smart contracts?


A) Reducing fees
B) Making them shorter
C) Encrypting user data
D) Detecting bugs and vulnerabilities

19. What is the limitation of DAI in blockchain?


A) Too creative
B) Needs large labeled data
C) Generates smart contracts
D) Works only offline

20. Which model is useful for smart contract code generation?


A) GDM
B) GAN
C) VAE
D) LLM
Paper 2 MCQ (21-25)
21. What is the main purpose of using GDM in blockchain?
A) Store data
B) Improve mining speed
C) Optimize performance
D) Reduce power usage

22. What does GAI use to enhance privacy?


A) Delete logs
B) Generate fake data
C) Encrypt all data
D) Reduce sharing

23. Generate synthetic data from real network traffic log to help training attack
detection model.
A) GAN
B) LLM
C) GDM
D) VAE

24. Which generative AI model uses an encoder to map input data into a latent space
and a decoder to generate samples?
A) GAN
B) GDM
C) LLM
D) VAE

25. Convert secret key to image and learn to generate secret shares based on images.
A) LLM
B) GDM
C) GAN
D) VAE
Paper 2 Essay (26-30)

26. How does DAI support secure blockchain operations?

It uses past data to detect known threats and optimize tasks.

27. How can GAI enhance privacy in AI systems?

It generates synthetic data to hide real user information.

28. How do LLMs improve smart contract auditing?

They understand code context to detect hidden vulnerabilities.

29. How does GDM improve blockchain throughput?

By learning configurations that allow more transactions per second (TPS).

30. How does GDM help reduce latency in blockchain networks?

By adjusting parameters to speed up transaction confirmation.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy