wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Data Science Fundamentals Worksheet

Total questions: 91

Worksheet time: 46mins

Name
Class
Date
1.

What is the primary goal of data science?

a)

To collect as much data as possible

b)

To use data to gain insights and make decisions

c)

To replace human judgment entirely

d)

To create complex algorithms without application

2.

Which of the following is an example of unstructured data?

a)

A CSV file of customer purchases

b)

A SQL database table

c)

A tweet containing text and emojis

d)

An Excel spreadsheet with sales figures

3.

What does the "Venn Diagram of Data Science" include?

a)

Hacking skills, math/statistics, and domain expertise

b)

Coding, visualization, and cloud computing

c)

Data mining, AI, and robotics

d)

SQL, Python, and R

4.

Which type of data is "zip code" classified as?

a)

Quantitative continuous

b)

Quantitative discrete

c)

Qualitative nominal

d)

Qualitative ordinal

5.

What is the key assumption of the Central Limit Theorem?

a)

Data must be normally distributed

b)

Sample means will approximate normality as sample size increases

c)

Population variance must be known

d)

Samples must be biased

6.

In hypothesis testing, what does a p-value < 0.05 typically indicate?

a)

Accept the null hypothesis

b)

Reject the null hypothesis

7.

Which metric measures "how often a classifier is correct when it predicts positive"?

a)

Recall

b)

F1-score

c)

Precision

d)

Accuracy

8.

What does Occam’s Razor suggest in modeling?

a)

Use the most complex model available

b)

The simplest explanation is usually best

c)

Always prioritize high accuracy over interpretability

d)

Add more features to improve performance

9.

Which distance metric is calculated as √(Σ(x_i - y_i)²)?

a)

Manhattan distance

b)

Euclidean distance

c)

Cosine similarity

d)

Hamming distance

10.

What is the purpose of z-score normalization?

a)

To convert categorical data to numerical

b)

To scale data to have mean=0 and standard deviation=1

c)

To remove outliers permanently

d)

To reduce dimensionality

11.

Which of the following is NOT a requirement for a distance metric?

a)

Positivity (d(x,y) ≥ 0)

b)

Symmetry (d(x,y) = d(y,x))

c)

Triangle inequality

d)

Linear correlation

12.

What does a correlation coefficient of -0.9 indicate?

a)

Strong positive relationship

b)

Weak negative relationship

c)

Strong negative relationship

d)

No relationship

13.

Which plot is best for visualizing relationships between two numerical variables?

a)

Histogram

b)

Box plot

c)

Scatter plot

d)

Pie chart

14.

What is the main weakness of pie charts?

a)

They are hard to read and compare

b)

They only work for small datasets

c)

They require 3D rendering

d)

They cannot show proportions

15.

Which rule states that ~68% of data falls within ±1 standard deviation of the mean?

a)

Bayes’ Theorem

b)

Law of Large Numbers

c)

Empirical Rule

d)

Central Limit Theorem

16.

In a confusion matrix, what does "False Negative" represent?

a)

Correctly predicted negative cases

b)

Incorrectly predicted positive cases

c)

Correctly predicted positive cases

d)

Incorrectly predicted negative cases

17.

The harmonic mean of precision and recall is called:

a)

F1 score

b)

Accuracy

c)

Specificity

d)

ROC curve

18.

Which sampling method ensures every member has an equal chance of selection?

a)

Stratified sampling

b)

Random sampling

c)

Convenience sampling

d)

Snowball sampling

19.

What is the purpose of the "bias-variance tradeoff"?

a)

To maximize model complexity

b)

To balance underfitting and overfitting

c)

To eliminate all errors

d)

To ignore outliers

20.

Which of the following is a categorical variable?

a)

Temperature in Celsius

b)

Height in cm

c)

Blood type (A/B/AB/O)

d)

Income in dollars

21.

What does a high coefficient of variation (CV) indicate?

a)

Low variability relative to the mean

b)

High variability relative to the mean

c)

Perfect correlation

d)

Normal distribution

22.

Which test checks if two categorical variables are independent?

a)

t-test

b)

ANOVA

c)

Chi-square test

d)

Kolmogorov-Smirnov test

23.

What is the primary use of a box plot?

a)

To show trends over time

b)

To visualize the distribution and outliers

c)

To compare proportions

d)

To display correlations

24.

Which of the following is a supervised learning algorithm?

a)

K-means clustering

b)

Logistic regression

c)

Principal Component Analysis (PCA)

d)

Apriori algorithm

25.

What does "imputation" refer to in data cleaning?

a)

Removing missing values

b)

Estimating missing values

c)

Standardizing scales

d)

Encoding categories

26.

What is the primary goal of logistic regression?

a)

To predict continuous outcomes

b)

To classify data into discrete categories

c)

To reduce dimensionality

d)

To cluster unlabeled data

27.

Which activation function is commonly used in logistic regression?

a)

ReLU

b)

Sigmoid

c)

Tanh

d)

Linear

28.

Which of the following is the main purpose of L2 regularization?

a)

Increase model complexity

b)

Reduce overfitting by penalizing large coefficients

c)

Ignore feature correlations

d)

Maximize training accuracy

29.

In K-means clustering, what does "K" represent?

a)

Number of features

b)

Number of clusters

c)

Learning rate

d)

Iteration count

30.

Which evaluation metric is most appropriate for imbalanced datasets?

a)

Accuracy

b)

F1-score

c)

Mean squared error (MSE)

d)

R-squared

31.

What is the key assumption of Naive Bayes classifiers?

a)

Features are normally distributed

b)

Features are conditionally independent

c)

Data must be linearly separable

d)

Labels are ordinal

32.

Which algorithm seeks the "maximum margin" hyperplane in classification?

a)

Decision Trees

b)

Support Vector Machines (SVM)

c)

K-nearest neighbors

d)

Random Forest

33.

What does "bagging" in ensemble methods refer to?

a)

Combining models with weighted votes

b)

Training models on random subsets of data with replacement

c)

Iteratively reweighting misclassified samples

d)

Reducing features via PCA

34.

Which of the following is a disadvantage of decision trees?

a)

High interpretability

b)

Prone to overfitting

c)

Handles only numerical data

d)

Requires feature scaling

35.

What is the purpose of the ROC curve?

a)

To visualize precision-recall tradeoffs

b)

To compare classifiers at different thresholds

c)

To measure clustering compactness

d)

To reduce feature dimensions

36.

Which technique helps address class imbalance?

a)

Removing minority class samples

b)

Oversampling the minority class

c)

Ignoring class weights

d)

Using Euclidean distance exclusively

37.

What does "gradient descent" optimize in linear regression?

a)

Sum of absolute errors

b)

Sum of squared errors

c)

Number of features

d)

Correlation coefficient

38.

Which statement about the "bias-variance tradeoff" is true?

a)

High bias implies overfitting

b)

High variance implies underfitting

c)

Increasing model complexity reduces bias but may increase variance

d)

Simple models always have low variance

39.

The main advantage of non-parametric models is:

a)

They do not assume a specific form for the underlying data distribution.

b)

They require fewer data points than parametric models.

c)

They are always faster to train than parametric models.

d)

They guarantee higher accuracy than parametric models.

40.

Which visualization is best for showing parts of a whole?

a)

Scatter plot

b)

Stacked bar chart

c)

Histogram

d)

Box plot

41.

What does a high "silhouette score" indicate in clustering?

a)

Poor separation between clusters

b)

Well-defined clusters

c)

Overfitting

d)

Random clustering

42.

Which method is used for dimensionality reduction?

a)

LASSO regression

b)

Principal Component Analysis (PCA)

c)

K-means clustering

d)

Logistic regression

43.

What is the key difference between supervised and unsupervised learning?

a)

Supervised learning uses labeled data

b)

Unsupervised learning always has higher accuracy

c)

Supervised learning ignores features

d)

Unsupervised learning requires a test set

44.

Which of the following is a hyperparameter?

a)

Coefficient in linear regression

b)

Learning rate in gradient descent

c)

Mean of the dataset

d)

Output prediction

45.

What problem does "multicollinearity" cause in regression?

a)

Reduces model interpretability

b)

Makes coefficient estimates unstable

c)

Increases training speed

d)

Eliminates outliers

46.

Which statement about "deep learning" is true?

a)

Requires minimal data

b)

Automatically performs feature engineering

c)

Only works with tabular data

d)

Avoids non-linear transformations

47.

What is the purpose of a "validation set"?

a)

To test the final model

b)

To tune hyperparameters

c)

To replace the training set

d)

To remove outliers

48.

Which metric is used to evaluate regression models?

a)

Precision

b)

Recall

c)

Mean Absolute Error (MAE)

d)

F1-score

49.

What does "one-hot encoding" do?

a)

Converts categorical variables to binary vectors

b)

Normalizes numerical data

c)

Reduces dimensionality

d)

Clusters similar features

50.

50. The algorithm that is most sensitive to feature scaling is:

a)

K-Nearest Neighbors

b)

Decision Tree

c)

Naive Bayes

d)

Random Forest

51.

Which of the following is an unsupervised learning algorithm?

a)

Decision Trees

b)

Random Forest

c)

K-means clustering

d)

Naive Bayes

52.

In Bayesian statistics, what does the "prior probability" represent?

a)

The probability after observing data

b)

The initial belief before observing data

c)

The likelihood of the null hypothesis

d)

The error rate of the model

53.

Which distribution is commonly used to model binary outcomes?

a)

Normal distribution

b)

Poisson distribution

c)

Bernoulli distribution

d)

Uniform distribution

54.

What is the key idea behind the "kernel trick" in SVM?

a)

It reduces computational complexity

b)

It transforms data into higher dimensions without explicit calculation

c)

It eliminates the need for feature scaling

d)

It works only with linear decision boundaries

55.

Which of the following is NOT a loss function for regression?

a)

Mean Squared Error (MSE)

b)

Cross-entropy

c)

Mean Absolute Error (MAE)

d)

Huber loss

56.

What does "entropy" measure in decision trees?

a)

The purity of a node

b)

The depth of the tree

57.

Which statement about "AdaBoost" is true?

a)

It gives equal weight to all classifiers

b)

It focuses on misclassified samples in subsequent iterations

c)

It requires features to be uncorrelated

d)

It only works with deep neural networks

58.

What is the primary use of the Kolmogorov-Smirnov test?

a)

To compare two distributions

b)

To evaluate classifier accuracy

c)

To reduce dimensionality

d)

To impute missing values

59.

Which probability rule states P(A|B) = P(B|A)P(A)/P(B)?

a)

Law of Total Probability

b)

Bayes’ Theorem

c)

Central Limit Theorem

d)

Empirical Rule

60.

What is the main challenge of "curse of dimensionality"?

a)

Increased computational cost and sparsity of data

b)

Decreased model interpretability

c)

Overfitting due to small sample sizes

d)

All of the above

61.

Which technique is used to handle overfitting in decision trees?

a)

Pruning

b)

Increasing tree depth

c)

Removing features

d)

Using larger datasets only

62.

In hypothesis testing, what is a Type I error?

a)

Failing to reject a false null hypothesis

b)

Rejecting a true null hypothesis

c)

Accepting the alternative hypothesis

d)

Incorrectly calculating the p-value

63.

Which of the following is a non-parametric test?

a)

t-test

b)

ANOVA

c)

Wilcoxon rank-sum test

d)

Chi-square test

64.

What does "AUC-ROC" measure?

a)

The area under the precision-recall curve

b)

The model’s ability to distinguish between classes

c)

The variance of the residuals

d)

The clustering compactness

65.

Which statement about "feature engineering" is true?

a)

It is unnecessary for deep learning models

b)

It involves creating new features from raw data

c)

It only applies to numerical data

d)

It reduces model accuracy

66.

What is the purpose of "stratified sampling"?

a)

To ensure each subgroup is proportionally represented

b)

To randomly shuffle data

c)

To remove outliers

d)

To reduce feature dimensions

67.

Which of the following is a disadvantage of KNN?

a)

High computational cost for large datasets

b)

Assumes linear separability

68.

What does "multiclass classification" involve?

a)

Predicting more than two classes

b)

Using multiple binary classifiers

c)

Both A and B

d)

None of the above

69.

Which distribution describes the number of events in fixed intervals?

a)

Normal distribution

b)

Binomial distribution

c)

Poisson distribution

d)

Exponential distribution

70.

What is the key advantage of "random forests" over single decision trees?

a)

Reduced overfitting through ensemble averaging

b)

Faster training time

c)

Lower memory usage

d)

No hyperparameters to tune

71.

Which metric is most sensitive to outliers in regression?

a)

Mean Absolute Error (MAE)

b)

Mean Squared Error (MSE)

c)

R-squared

d)

F1-score

72.

What is the primary goal of "dimensionality reduction"?

a)

To remove all categorical features

b)

To retain important information while reducing features

c)

To increase model complexity

d)

To eliminate the need for training data

73.

Which statement about "cross-validation" is true?

a)

It reduces the need for a test set

b)

It provides a more robust estimate of model performance

c)

It is only used for clustering

d)

It requires labeled data in unsupervised learning

74.

What does "TF-IDF" stand for in text processing?

a)

Term Frequency-Inverse Document Frequency

b)

Text Feature-Indexed Data Format

c)

Time-Frequency Inverse Distribution

d)

Tokenized Frequency-Integrated Data Frame

75.

Which algorithm is used for anomaly detection?

a)

Linear Regression

b)

K-means

c)

Isolation Forest

d)

Logistic Regression

76.

What is the main purpose of "data augmentation"?

a)

To delete redundant data

b)

To artificially increase training data diversity

c)

To reduce dataset size

d)

To replace missing values

77.

What does a z-score of 2.0 indicate?

a)

The value is 2 standard deviations below the mean

b)

The value is 2 standard deviations above the mean

c)

The value is in the bottom 5% of the distribution

d)

The value is exactly at the mean

78.

Which R function correctly calculates the mean of a vector x?

a)

median(x)

b)

mean(x)

79.

The Naive Bayes classifier assumes:

a)

Features are normally distributed

b)

Features are conditionally independent given the class

c)

All features are equally important

d)

The data has no missing values

80.

In hypothesis testing, what does the p-value represent?

a)

The probability that the null hypothesis is true

b)

The probability of observing the data if the null hypothesis is true

c)

The effect size of the alternative hypothesis

d)

The confidence level

81.

Which R code would you use to remove rows with missing values from a dataframe df?

a)

df[complete.cases(df), ]

b)

df %>% drop_na()

c)

na.omit(df)

d)

All of the above

82.

The Central Limit Theorem states that:

a)

Sample means are normally distributed if n ≥ 30

b)

Population distributions must be normal

c)

Variance decreases with larger samples

d)

Confidence intervals always contain the true mean

83.

Which metric is most appropriate for evaluating an imbalanced binary classifier?

a)

Accuracy

b)

F1-score

c)

R-squared

d)

Euclidean distance

84.

What does the empirical rule state about normally distributed data?

a)

68% falls within ±1 SD, 95% within ±2 SD

b)

The mean equals the median

c)

Skewness must be zero

d)

It only applies to large samples

85.

In decision trees, how is the optimal split chosen?

a)

By maximizing information gain

b)

By minimizing Gini impurity

c)

Both A and B

d)

By random selection

86.

Which R package is used for creating heatmaps?

a)

ggplot2

b)

dplyr

c)

heatmap() (base R)

d)

Both A and C

87.

The Kolmogorov-Smirnov test compares:

a)

Two distributions

b)

Means of paired samples

c)

Variance across groups

d)

Correlation coefficients

88.

What does mutate() do in R (dplyr)?

a)

Adds new variables to a dataframe

b)

Filters rows

c)

Sorts columns

d)

Merges tables

89.

Which of the following formulas is correct?

a)

TP / (TP + FP)

b)

TP / (TP + FN)

c)

(TP + TN) / total

d)

FP / (FP + TN)

90.

Which statement about SVM is true?

a)

It only works with linear decision boundaries

b)

It maximizes the margin between classes

c)

It ignores outliers

d)

It requires features to be uncorrelated

91.

What does this R code do?

a)

Filters high salaries

b)

Calculates average salary by department

c)

Creates a scatter plot

d)

Removes missing values