NPTEL Edge Computing Week 4 Assignment Answers 2025
1. Which algorithm is used to aggregate models in Federated Learning?
- Backpropagation
- FedAvg
- K-means
- Principal Component Analysis
Answer :- For Answers Click Here
2. What is one major advantage of Federated Learning for IoT devices?
- Reduced hardware requirements
- Simplified data preprocessing
- Data privacy and decentralized training
- Increased computational speed
Answer :-
3. In knowledge distillation, which loss function is used to transfer knowledge from teacher to student?
- Mean Squared Error (MSE)
- Hinge Loss
- Cross-Entropy
- KL Divergence
Answer :-
4. What is the primary challenge of non-IID (independent and identically distributed) data in FL?
- Data leakage
- Higher computational costs
- Slower convergence of the model
- Model overfitting
Answer :-
5. What is the primary benefit of low-rank factorization in deep learning model optimization?
- Reduces model size and computation costs
- Increases model accuracy
- Improves training speed
- Enhances data privacy
Answer :- For Answers Click Here
6. In Federated Learning, if a cluster has 4 devices and each device achieves an accuracy of 80%, what is the average accuracy
of the aggregated model?
- 20%
- 40%
- 100%
- 80%
Answer :-
7. What is a critical privacy concern in Federated Learning?
- Membership inference attacks
- Model convergence speed
- Computational power
- Network bandwidth
Answer :-
8. A Federated Learning model sends updates to the central server every 5 seconds. How many updates are sent in 1 minute?
- 10
- 12
- 15
- 20
Answer :-
9. In Knowledge Distillation, which of the following is considered the “teacher”?
- A smaller, simplified model
- The dataset used for training
- A large, pre-trained model
- The optimizer used during training
Answer :-
10. What challenge in Federated Learning does the term “non-IID data” refer to?
- High bandwidth usage during model training
- Data being unevenly distributed and diverse across devices
- Lack of encryption for sensitive data
- Overfitting due to excessive training
Answer :- For Answers Click Here
