NPTEL Business Intelligence & Analytics Week 6 Assignment Answers 2025
1.
Answer :- For Answers Click Here
2. During a cricket match, Hotstar’s “Selfie with an Old Cricket Bat” campaign targets 30% of college students and achieves an 8% subscription rate. Predictive analytics targeting the top 30% achieves a 16% subscription rate. What is the lift achieved?
- 1.5
- 2.0
- 2.5
- 3.0
Answer :-
3. In a study on water quality, a classifier predicts whether water is contaminated. Here’s the confusion matrix for contaminated = yes and contaminated = no.
What is the specificity of the classifier?
- 96.84%
- 97.12%
- 98.01%
- 97.92%
Answer :-
4. A car manufacturer wants to predict customer satisfaction based on survey data, but the responses are noisy and have irrelevant details. How does this noise affect the prediction model?
- It increases the size of the dataset
- It increases the complexity of the model
- It provides irrelevant information to the model
- It improves the accuracy of predictions
Answer :-
5. A movie recommendation platform suggests a lot of movies to its users, but many of the recommended movies are not relevant to the user’s genre taste. This would represent:
- High recall, Low precision
- High recall, High precision
- Low recall, High precision
- Low recall, Low precision
Answer :-
6. Which technique is known to enhance the accuracy of a classifier trained on imbalanced data without modifying the base model?
- Random undersampling
- Ensemble techniques
- Data augmentation
- Model-specific cost-sensitive learning
Answer :-
7. Identify two models that are considered classic linear classifiers:
- Logistic regression, K-Nearest Neighbors
- Decision tree, Random forest
- Perceptron, Logistic regression
- Decision tree, Neural networks
Answer :-
8. Which aspect does the F-measure focus on in the evaluation of classification models?
- The computational speed during prediction
- The balance between precision and recall
- The simplicity of the decision boundary
- The accuracy of the confusion matrix
Answer :-
9. In a weather forecasting system, RMSE is used to assess the model’s accuracy. What does it measure?
- The computational resources required for prediction
- The number of days included in the forecast dataset
- The relationship between temperature and humidity variables
- The average error in predicted and observed weather conditions
Answer :-
10. To ensure an equal number of positive and negative tuples in the training dataset, which technique involves replicating the positive tuples?
- Threshold moving
- Ensemble techniques
- Undersampling
- Oversampling
Answer :- For Answers Click Here
11. Which of the following is part of the learning step in classification?
- Constructing a classification model
- Applying the model to test data
- Estimating the model’s accuracy
- Evaluating the model’s performance
Answer :-
12. Using the provided ROC curves for models M1 and M2, identify the model with higher accuracy.
- Both models are equally accurate
- M1
- Accuracy cannot be assessed from this curve
- M2
Answer :-
13. The term “reducible error” in classification refers to:
- Error that can be reduced through improved learning approaches
- Error caused by the data collection process
- Error due to uncontrollable randomness
- Error that remains fixed regardless of the model used
Answer :-
14. What does the formula (FP+FN) / (P+N) represent in classification metrics?
- Specificity
- Sensitivity
- Accuracy
- Error rate
Answer :-
15. Identify the technique that is NOT used for classification:
- Logistic Regression
- Quadratic Discriminant Analysis
- Random Forest
- DBSCAN
Answer :- For Answers Click Here
