Question bank
Pick an answer to see if you're right — with an explanation.
All topicsPythonSQL / MySQLData ScienceNumPyPandasMatplotlibMachine LearningDeep LearningArtificial IntelligenceLarge Language ModelsDjangoAPIsGit & GitHubPower BIExcelStatistics
Data ScienceEasy
Which is a leading indicator for monthly revenue in an online store?
aLast quarter's profit
bAnnual churn rate
cAudited yearly turnover
dItems added to cart this week
Explanation: Cart additions move before revenue does, giving time to react; the others are lagging.
Deep LearningEasy
Which layer type is best suited to capturing local spatial patterns in images?
ann.LSTM
bnn.Embedding
cnn.Conv2d
dnn.LayerNorm
Explanation: Convolutions slide shared filters over the image, detecting local patterns like edges anywhere they occur.
DjangoEasy
Which decorator redirects anonymous users to the login page?
a@login_required
b@csrf_protect
c@require_POST
d@never_cache
Explanation: `@login_required` sends unauthenticated users to `LOGIN_URL`.
ExcelEasy
Which shortcut triggers Flash Fill?
aCtrl+T
bCtrl+Shift+L
cCtrl+D
dCtrl+E
Explanation: Ctrl+E runs Flash Fill, which completes a column from the pattern you typed in the first cell.
Git & GitHubEasy
A commit already pushed to `main` introduced a bug. What is the safest way to undo it?
agit reset --hard HEAD~1 and force-push
bgit revert <sha>
cDelete the repository
dgit commit --amend
Explanation: revert adds a new commit that undoes the change without rewriting shared history.
Large Language ModelsEasy
Why do Hindi or Tamil prompts often cost more than English prompts of the same meaning?
aProviders charge a language surcharge
bThey are sent twice
cThey usually split into more tokens
dThey require a larger model
Explanation: Tokenizers trained mostly on English break many Indic-script words into more pieces.
Machine LearningEasy
Which scikit-learn model tends to set some coefficients exactly to zero?
aLasso
bRidge
cLinearRegression
dKNeighborsRegressor
Explanation: The L1 penalty in Lasso can shrink weights all the way to zero, acting as feature selection.
MatplotlibEasy
Which call is the recommended way to create a figure and one Axes in object-oriented style?
aplt.figure().plot()
bplt.axes_create()
cfig, ax = plt.subplots()
dax = plt.Axes()
Explanation: `plt.subplots()` returns the Figure and Axes so you can call methods on them explicitly.
NumPyEasy
Which is the recommended way to create random numbers in modern NumPy?
arng = np.random.default_rng(42)
bnp.random.seed(42); np.random.rand()
cimport random; random.random()
dnp.rand(42)
Explanation: The Generator API from default_rng is the modern, recommended interface.
PandasEasy
What does `pd.read_excel('file.xlsx', sheet_name=None)` return?
aOnly the first sheet
bA list of column names
cA dict mapping sheet names to DataFrames
dThe number of sheets
Explanation: sheet_name=None reads every sheet into a dictionary keyed by sheet name.
Power BIEasy
You have three depot files with identical columns and want a single table containing all their rows. Which operation do you use?
aMerge Queries
bGroup By
cAppend Queries
dUnpivot Columns
Explanation: Append is the UNION ALL of Power Query — it stacks rows; Merge adds columns via a join.
PythonEasy
Which module provides `Counter` and `defaultdict`?
aitertools
bfunctools
ccollections
dstatistics
Explanation: Both live in the `collections` module of the standard library.
SQL / MySQLEasy
Which statement permanently saves the changes made in a transaction?
aSAVE
bROLLBACK
cCOMMIT
dEND
Explanation: COMMIT makes the transaction's changes durable; ROLLBACK discards them.
StatisticsEasy
If you quadruple the sample size, what happens to the standard error of the mean?
aIt halves
bIt quarters
cIt doubles
dIt stays the same
Explanation: SE = sigma / sqrt(n); sqrt(4) = 2, so the standard error halves.
PythonEasy
What does this print?
x = [1, 2, 3]
print(x[-1])a3
b1
cError
d-1
Explanation: Negative indexes count from the end.