Indian Institute of Technology Madras
BS Degree Programme · Examination & Study Paper
Tools in Data Science
Regression, ML & time series
Examination Instructions & Guidelines
- Solve on notebook/paper first before checking solutions.
- MCQ: Exactly 1 option is correct. Negative marking applies.
- Scientific calculator allowed for all calculations.
- Official answer key is sealed at the end of this sheet.
End Term · 13 Apr 2025 (FN)End Term
- Q1MCQ1 Mark·Choose one option
A pharmaceutical researcher is testing a new drug's effectiveness. If the p-value is 0.03 in a regression analysis, what can be concluded about the drug's impact?
- AThe drug has no significant effect
- BThe result is statistically significant at a 5% significance level
- CMore data collection is needed
- DThe sample size is too small
End Term · 22 Dec 2024 (FN)End Term
- Q2MCQ1 Mark·Choose one option
You and your friend are building a KNN model to classify customers based on their shopping patterns. You use cross-validation to choose K = 6, while your friend guesses and sets K = 2. On the test set, you notice that your friend's model has a much higher training accuracy but much lower test accuracy than yours. Which of the following is the best explanation?
- AYour friend’s model has higher bias than yours
- BYour friend’s model has higher variance and overfits the data
- CYour model underfits the data compared to your friend’s model
- DYour model overfits the data compared to your friend’s model
- Q3MCQ1 Mark·Choose one option
You are running a K-means clustering algorithm on customer segmentation data. Each time you run the model, you get different cluster assignments, making it hard to compare results. Which parameter should you adjust in the kmeans() function to achieve more stable and consistent results?
- Aalgorithm
- Bmax_iter
- Cn_clusters
- Dn_init
- Q4MCQ1 Mark·Choose one option
You are fitting an ARIMA model to a stock price time series. You want to use no moving average component, 5 lags for autoregression, and 1 difference to make the series stationary. Which of the following model specifications should you use?
- AARIMA(..., trend = (5,1,0))
- BARIMA(..., order = (5,1,0))
- CARIMA(..., order = (0,5,1))
- DARIMA(..., trend = (0,5,1))
- Q5MCQ1 Mark·Choose one option
You are building a linear regression model to predict house prices based on the number of bedrooms. You obtain a p-value of 0.03 for the coefficient of "number of bedrooms." Which of the following conclusions can you draw?
- AThe relationship between number of bedrooms and house prices is statistically significant
- BThere is no relationship between number of bedrooms and house prices
- CThe analysis is inconclusive
- DThe model is overfitting