UGC-NET Business Analytics & Digital Agribusiness: practice questions

1 · Supervised learning: regression · medium

A demand model has errors of 4, −2, 6 and −8 tonnes on four test weeks. What is the root mean squared error (RMSE)?

1. 5.48
2. 5.00
3. 0.00
4. 30.00

2 · Supervised learning: regression · medium

In a logistic regression of technology adoption, the coefficient on 'member of a farmer producer organisation' (1 = yes) is 0.693. How is it interpreted?

1. Members have a 0.693 probability of adopting, regardless of their other traits
2. Members are 69.3 percentage points more likely to adopt than non-members
3. Members have twice the odds of adopting compared with non-members
4. Membership explains 69.3% of the variation in adoption across the sample farms

3 · Supervised learning: regression · medium

A crop-disease classifier is tested on 200 fields: 40 true positives, 10 false positives, 20 false negatives and 130 true negatives. What is its recall (sensitivity)?

1. 0.80
2. 0.67
3. 0.85
4. 0.93

4 · Supervised learning: regression · medium

A crop-disease classifier is tested on 200 fields: 40 true positives, 10 false positives, 20 false negatives and 130 true negatives. What is its F1 score?

1. 0.73
2. 0.85
3. 0.53
4. 0.67

5 · Supervised learning: regression · medium

To test a price-forecasting model for onions on six years of weekly data, which evaluation design is correct?

1. Fit on the first five years and test on the sixth, moving forward in time
2. Fit on the sixth year and test on the first five years to obtain a larger test set
3. Fit and test on all six years so that every seasonal pattern appears in both sets
4. Randomly select 20% of weeks from all years as a test set and fit on the rest

6 · Linear discriminant analysis · medium

When is logistic regression usually preferred over linear discriminant analysis for a two-group problem?

1. When the groups share equal covariance matrices and the predictors are normal
2. When predictors include dummy variables and are clearly not multivariate normal
3. When there are more than two groups and a low-dimensional plot of them is wanted
4. When the sample is small and the most statistically efficient estimates are needed

7 · Linear discriminant analysis · hard

In discriminant output, why do analysts often interpret the structure matrix (loadings) rather than the standardised coefficients when predictors are correlated?

1. The structure matrix is the only output that remains valid when the groups have unequal sizes
2. Correlated predictors make coefficients unstable or flip sign; loadings are plain correlations
3. Standardised coefficients are expressed in the original units and cannot be compared at all
4. Loadings always sum to one, so they show each predictor's exact share of discrimination

8 · Digital marketing and brand building · medium

A food brand spent 50,000 on ads that generated 180,000 of attributed revenue at a 20% gross margin. Which statement is correct?

1. ROAS is 3.6, yet the ads lost money because gross profit of 36,000 is below the spend
2. ROAS is 3.6, so the campaign earned a net profit of 130,000 after paying for all the ads
3. ROAS is 0.28, because ad spend is divided by the revenue the ads generated
4. ROAS is 0.72, because return on ad spend is always calculated from gross profit

9 · Digital marketing and brand building · hard

An animal-feed brand earns an annual margin of 1,200 per retained dairy customer. Annual retention is 70% and the discount rate 10%. Using CLV = m × r/(1 + d − r), what is customer lifetime value?

1. 4,000
2. 2,100
3. 3,000
4. 840

10 · Digital marketing and brand building · hard

A team wants to detect an increase in conversion from 5% to 6% with a two-sided 5% test and 80% power. Using n = (1.96 + 0.84)² × [p1(1 − p1) + p2(1 − p2)]/(p2 − p1)², about how many visitors are needed per version?

1. 81,600
2. 82
3. 816
4. 8,200