NPTEL Business Intelligence & Analytics Week 11 Assignment Answers 2025
1. Which type of layer is NOT a standard component of a Convolutional Neural Network (CNN)?
- Convolutional layer
- Recurrent layer
- Pooling layer
- Fully Connected (FC) layer
Answer :- For Answers Click Here
2. Which of the following is a potential source of bias in AI models?
- Training data bias
- Algorithmic bias
- Cognitive bias
- All of the above
Answer :-
3. In a perceptron, after calculating the weighted sum of inputs, the _ function is applied to determine the final neuron output
- Loss
- Transfer
- Input
- Combination
Answer :-
4. Which characteristic differentiates a feedforward neural network (FNN) from RNNs and CNNs?
- FNN has feedback connections for learning sequences
- FNN processes data using convolutional filters.
- FNN has no cycles, and information flows in only one direction.
- FNN updates weights dynamically based on past outputs.
Answer :-
5. A researcher is studying the impact of salary changes on employee productivity over 5 years, collecting data from the same employees annually. What type of data is being used?
- Cross-sectional data
- Time-series data
- Panel data
- Experimental data
Answer :-
6. ___________ helps explain the decision-making process of AI models, making them more transparent and trustworthy for human users.
- Deep Learning
- Explainable AI (XAI)
- Artificial General Intelligence (AGI)
- Neural Networks
Answer :-
7. What does the trend component in time series analysis represent?
- Short-term fluctuations in the data
- Repeating patterns that occur within a fixed period
- Long-term movement or general direction of the data
- Random variations with no specific pattern
Answer :- For Answers Click Here
8. A neural network has 14 input neurons and 6 output neurons. The recommended number of hidden neurons using the thumb rule is __.
- 5
- 10
- 14
- 20
Answer :-
9. In blockchain, transactions are validated after they are linked into blocks.
- True
- False
Answer :-
10. In responsible AI, the principle of __ ensures that machine learning models do not create biased outcomes against protected groups based on race, gender, or socioeconomic status.
- Privacy
- Transparency
- Fairness
- Explainability
Answer :-
11. Which neural network layer is responsible for capturing complex patterns and nonlinear relationships in data?
- Input layer
- Output layer
- Hidden layer
- Loss layer
Answer :-
12. An __ function is used in neural networks to introduce non-linearity, allowing the model to learn complex patterns from data.
- Loss
- Activation
- Combination
- Optimization
Answer :-
13. Which of the following machine learning models provides the highest level of interpretability and is easiest for humans to understand?
- Random Forest
- Neural Networks
- Decision Tree
- Support Vector Machine (SVM)
Answer :-
14. Which mechanism guarantees that once a transaction is recorded on a blockchain, it cannot be altered?
- Centralized control
- Smart contracts
- Encryption
- Cryptographic hashing
Answer :-
15. Recurrent Neural Networks (RNNs) use _ to compute gradients and update weights over sequential data.
- Gradient Clipping
- Backpropagation Through Time (BPTT)
- Adam Optimizer
- Dropout Regularization
Answer :- For Answers Click Here
