GATE - Data Science and Artificial Intelligence (DA) · India · IITs and IISc on rotation (2027: IIT Madras)
Original practice questions written to the published syllabus · not past papersGATE - Data Science and Artificial Intelligence (DA) is a postgraduate admission exam run by IITs and IISc on rotation (2027: IIT Madras), for candidates in India. Computer-based, 3 hours, 100 marks (15 General Aptitude + 85 subject). MCQ, MSQ and NAT; negative marking on MCQ only.
Source: gate2027.iitm.ac.in.
Checked against the exam body's published information. Formats and fees change; confirm the current edition there before you book.
Looking at data before modelling it: summaries, robust statistics, graphs that show distributions and relationships honestly, outlier detection, and the tidy-data and reporting habits that make an analysis checkable. Deliberately practical and software-based.
Our 54 GATE DA questions in this subject: 12 easy, 29 medium, 13 hard, across 9 units.
Choose an answer, then open the working. Each explanation says why the right answer is right and why the others are not.
These 10 were written to GATE DA and are not shown on any other exam's page.
For which variable is the coefficient of variation a meaningful summary?
A. Milk yield per cow in litres per day
The CV divides the SD by the mean, which is meaningful only on a ratio scale with a true zero. With Celsius or dates, moving the zero changes the CV arbitrarily; codes for soil series are nominal.
A sample of 1,000 plant heights has IQR 4.0 cm. What bin width does the Freedman-Diaconis rule suggest?
A. 0.80 cm
h = 2 IQR n^(−1/3) = 2 × 4.0 × 1000^(−1/3) = 8.0/10 = 0.80 cm. 8.00 omits the n^(−1/3) factor; 0.40 omits the factor 2.
Silverman's rule of thumb is h = 0.9 min(s, IQR/1.34) n^(−1/5). For n = 256, s = 2.0 and IQR = 2.4, what is h?
B. 0.532
IQR/1.34 = 1.791 < 2.0, so min = 1.791; 256^(−1/5) = 0.3299; h = 0.9 × 1.791 × 0.3299 = 0.532. Using s instead of the minimum gives 0.594.
A kernel density estimate of germination percentages shows positive density below 0% and above 100%. What causes this?
B. Kernels near a natural boundary leak probability across it
Each kernel spreads mass on both sides of its observation, so values near 0 or 100 spill over the boundary. Boundary corrections or estimating on a transformed scale (logit) avoid it.
In a sample of 5 observations, what is the largest possible absolute z-score, (x − x̄)/s, using s with divisor n − 1?
B. 1.79
The bound is (n − 1)/√n = 4/√5 = 1.79. So in small samples a |z| > 3 rule can never flag anything: an extreme value inflates the s that defines its own z (masking). 2.24 is √5.
Seven yields have median 3.4 t/ha and MAD 0.2 t/ha. What is the modified z-score of a yield of 8.9 t/ha?
C. 18.55
Mᵢ = 0.6745 (8.9 − 3.4)/0.2 = 0.6745 × 27.5 = 18.55, far above the 3.5 flag. 27.50 omits the 0.6745 constant; 3.50 is the flagging threshold itself.
What is the Box-Cox transformation when λ = 0?
D. log y
y(λ) = (y^λ − 1)/λ tends to log y as λ → 0, which is why the log occupies the zero place on the ladder.
According to Cleveland and McGill, which visual encoding is read most accurately?
B. Position on a common scale
Their ranking runs from position on a common scale, through non-aligned position, length, angle and area, to volume and colour. This is why dot and bar charts beat pie and bubble charts.
Which chart is least accurate for comparing the shares of 9 crops that each occupy 9% to 14% of a region's area?
D. A pie chart
Small differences in angle and area are hard to judge. A sorted dot or bar chart puts the shares on a common scale, where differences of a percentage point are visible.
When is a table preferable to a graph?
D. When readers need to look up exact values
Tables serve look-up of precise numbers; graphs show patterns, trends, shapes and relations better. Many reports pair a graph for the message with a table for the record.
This plan follows where our questions actually fall, so the time goes where the marks are likely to be. It is our reading of the syllabus, not the exam body's weighting.
Read the manual chapters for these first and do every worked question you can find on them.
Aim for one sitting per unit, then mixed questions across them.
Learn the definitions, classifications and standard numbers; these units are more often recalled than worked.
Come back to these after a first pass, untimed, and read every explanation, including for the options you rejected.
Do these early in revision so they are fast on the day.
Last updated · How Laisha writes and checks its pages · Report a mistake
US$10 a month · one price worldwide · cancel any time
| Free | Smarter | |
|---|---|---|
| Grow guides, pest pages, articles | Read all | Read all |
| Ask Laisha | 5 a day | Up to 1,000 a month |
| Exam practice | The sample on each page | Every question, timed papers |
| Crop calendar and pest alerts | — | For your own fields |
| Calculators and tools | — | All |
| Uploads and reports | — | 100 a month, real files |
An AI operating system for agriculture: free guides, manuals and tools written for every country.