Disclaimer:
Educational Use Only:
The website links provided are to be used strictly for educational and knowledge-gaining purposes only.
No Misuse:
Do not misuse the links or the information contained therein for any commercial, illegal, or unethical activities.
Respect for Websites and Businesses:
Respect the intellectual property and business interests of the websites and their owners. Do not attempt to alter, copy, or redistribute content without proper authorization.
Acknowledge Value of Knowledge:
Recognize and value the knowledge and resources shared by these websites. Use the information responsibly and ethically.
No Endorsement:
The provision of these links does not imply endorsement of the linked sites, their content, or their products and services. Users are responsible for evaluating the suitability and accuracy of the information.
User Responsibility:
By accessing these links, you agree to use the information at your own risk and to comply with all applicable laws and website terms of use.
Overall MCQs for Reference
- What is the primary goal of Principal Component Analysis (PCA)?
a. Classification
b. Clustering
c. Dimensionality reduction
d. Regression
Answer: c. Dimensionality reductionstudocu - In PCA, what are the principal components?
a. Features of the dataset
b. Eigenvalues of the covariance matrix
c. Eigenvectors of the covariance matrix
d. Random vectors
Answer: c. Eigenvectors of the covariance matrixstudocu - Which step is crucial before applying PCA for best results?
a. Normalizing the dataset
b. Standardizing the dataset
c. Sorting the data
d. Combining features
Answer: b. Standardizing the datasetgithub - What does PCA maximize in a dataset?
a. Correlation between features
b. Variance along new dimensions
c. Number of features
d. Data points
Answer: b. Variance along new dimensionsplacementpreparation - How does PCA transform the original features?
a. Into dependent variables
b. Into orthogonal components
c. By scaling them equally
d. By clustering the data
Answer: b. Into orthogonal componentsplacementpreparation - Which matrix is used in PCA to compute eigenvectors and eigenvalues?
a. Distance matrix
b. Correlation matrix
c. Covariance matrix
d. Identity matrix
Answer: c. Covariance matrixgithub - What does an eigenvalue represent in PCA?
a. Direction of principal component
b. Magnitude of the principal component’s variance
c. Number of features combined
d. Data mean
Answer: b. Magnitude of the principal component’s variancegithub - When is PCA generally used?
a. To increase the dimensionality of data
b. To reduce the number of features while retaining variance
c. To create clusters in labeled data
d. For regression analysis
Answer: b. To reduce the number of features while retaining variancegeeksforgeeks - What is the impact of PCA on the interpretability of features?
a. Increases interpretability
b. Keeps original features unchanged
c. New features are linear combinations, so interpretability may reduce
d. Eliminates noise directly
Answer: c. New features are linear combinations, so interpretability may reducestudocu - What is the explained variance ratio in PCA?
a. Ratio of original features to components
b. Explained variance by a principal component divided by total variance
c. Ratio of eigenvectors to eigenvalues
d. Proportion of data lost
Answer: b. Explained variance by a principal component divided by total variancegithub - Which Python function is used to implement PCA in sklearn?
a. PCA()
b. pca_reduction()
c. PrincipalComponent()
d. ComponentAnalysis()
Answer: a. PCA()placementpreparation - How does PCA treat highly correlated features?
a. Amplifies them
b. Removes them entirely
c. Combines them into principal components
d. Leaves them untouched
Answer: c. Combines them into principal componentsgeeksforgeeks - Which of the following is NOT a step in PCA?
a. Calculate covariance matrix
b. Compute eigenvectors and eigenvalues
c. Normalize principal components
d. Project data onto selected components
Answer: c. Normalize principal componentsgithub - PCA results in components that are:
a. Dependent variables
b. Correlated features
c. Orthogonal (uncorrelated) components
d. Clusters of data points
Answer: c. Orthogonal (uncorrelated) componentsgithub - How many principal components can PCA have at most?
a. Equal to the number of original features
b. Equal to the number of data points
c. Fixed at 3
d. Unlimited
Answer: a. Equal to the number of original featuresgatevidyalay - What is the role of centering the data in PCA?
a. To normalize variances
b. To ensure zero mean for each feature
c. To reduce noise
d. To increase data variance
Answer: b. To ensure zero mean for each featuregithub - True or False: PCA is sensitive to the scale of the features.
a. True
b. False
Answer: a. Trueseas.upenn - What kind of data problems can PCA help address?
a. Data imbalance
b. High dimensionality and multicollinearity
c. Low sample size
d. Missing values
Answer: b. High dimensionality and multicollinearitygeeksforgeeks - Can PCA be used for feature selection directly?
a. Yes, by selecting original features
b. No, PCA transforms features rather than selecting them
c. Yes, by removing features with low variance
d. Sometimes, if constrained
Answer: b. No, PCA transforms features rather than selecting themmlstack - What does the projection matrix in PCA consist of?
a. Top k eigenvectors
b. Row means
c. Eigenvalues
d. Data cluster centers
Answer: a. Top k eigenvectorsgithub - Which type of matrix decomposition can be used to solve PCA?
a. QR decomposition
b. Singular Value Decomposition (SVD)
c. Cholesky decomposition
d. LU decomposition
Answer: b. Singular Value Decomposition (SVD)seas.upenn - What is the effect of retaining fewer principal components?
a. More information preserved
b. Less data compression
c. Potential information loss but dimensionality reduction
d. Increased data noise
Answer: c. Potential information loss but dimensionality reductiongithub - In PCA, what is an eigenvector?
a. A scalar indicating variance
b. A direction in feature space indicating maximum variance
c. The mean vector
d. Data centroid
Answer: b. A direction in feature space indicating maximum variancegithub - Which of these is a disadvantage of PCA?
a. Preserves all original features intact
b. Can make features harder to interpret
c. Requires no preprocessing
d. Always improves model accuracy
Answer: b. Can make features harder to interpretgeeksforgeeks - How are the principal components ranked?
a. By magnitude of eigenvalues in descending order
b. By number of features combined
c. Alphabetically by feature name
d. Randomly
Answer: a. By magnitude of eigenvalues in descending ordergithub - In a 2D dataset, how many principal components can there be?
a. One
b. Two
c. Three
d. Unlimited
Answer: b. Twogatevidyalay - What is the covariance matrix used for in PCA?
a. Compute variance of single feature
b. Measure joint variability between pairs of features
c. To compute means
d. To cluster features
Answer: b. Measure joint variability between pairs of featuresgithub - What does the first principal component capture?
a. The smallest variance direction
b. The direction with maximum variance
c. Random noise
d. Average value of all features
Answer: b. The direction with maximum variancegithub - What happens to the feature space after PCA transformation?
a. It expands exponentially
b. It remains unchanged
c. It changes to a new orthogonal coordinate system
d. It becomes non-linear
Answer: c. It changes to a new orthogonal coordinate systemgithub - Is PCA a supervised or unsupervised technique?
a. Supervised
b. Unsupervised
c. Semi-supervised
d. Reinforcement learning
Answer: b. Unsupervisedgeeksforgeeks - Which technique is PCA most closely related to?
a. Clustering
b. Linear Discriminant Analysis (LDA)
c. Decision Trees
d. Neural Networks
Answer: b. Linear Discriminant Analysis (LDA)gatevidyalay - For PCA, what does it mean if the explained variance ratio for the first few components is high?
a. Most of the data variance can be represented in fewer components
b. Data is very noisy
c. PCA is not effective
d. Features are independent
Answer: a. Most of the data variance can be represented in fewer componentsgithub - What is the main reason to perform PCA before other learning algorithms?
a. To increase feature count
b. To reduce dimensionality and computational cost
c. To create clusters
d. To label data points
Answer: b. To reduce dimensionality and computational costgeeksforgeeks - How does PCA handle non-linear relationships in data?
a. Fully captures them
b. Ignores them as it is a linear method
c. Enhances them
d. Converts them to linear
Answer: b. Ignores them as it is a linear methodaimcqs - Which of the following is NOT true about PCA?
a. It can help visualize high dimensional data
b. It creates new features that are combinations of original features
c. It increases the dimensionality
d. It reduces redundancy in data
Answer: c. It increases the dimensionalitygeeksforgeeks - What is the meaning of orthogonality in PCA components?
a. Components are parallel
b. Components are independent and uncorrelated
c. Components are negatively correlated
d. Components are dependent on each other
Answer: b. Components are independent and uncorrelatedgithub - What is the typical first step in PCA on data matrix X?
a. Subtract mean from each feature
b. Scale each feature tostudocu
c. Remove outliers
d. Add noise for regularization
Answer: a. Subtract mean from each featuregithub - How does PCA contribute to noise reduction?
a. By keeping components associated with high variance and discarding others
b. By normalizing data
c. By clustering noise points
d. By increasing feature correlation
Answer: a. By keeping components associated with high variance and discarding othersgeeksforgeeks - PCA components are:
a. Always interpretable as original features
b. Unrelated to original features
c. Linear combinations of original features
d. Equal to original features
Answer: c. Linear combinations of original featuresstudocu - Which algorithm can PCA be combined with for improved visualization?
a. k-Nearest Neighbors
b. t-SNE
c. SVM
d. Random Forest
Answer: b. t-SNEplacementpreparation - What is the key assumption behind PCA?
a. Features are independent
b. Maximize variance along principal axes to represent data
c. Data is categorical
d. Classes are linearly separable
Answer: b. Maximize variance along principal axes to represent dataplacementpreparation - How do you choose the number of principal components to retain?
a. Fixed value of 2
b. Based on cumulative explained variance threshold (e.g., 95%)
c. Random selection
d. Equal to sample size
Answer: b. Based on cumulative explained variance threshold (e.g., 95%)github - True or False: PCA can be applied to any kind of data without preprocessing.
a. True
b. False
Answer: b. Falsegithub - Does PCA perform better on data with:
a. Mixed scales if unstandardized
b. Features on the same scale
c. Noisy data without preprocessing
d. Categorical variables only
Answer: b. Features on the same scalegithub - What kind of variance does PCA seek to preserve?
a. Noise variance
b. Maximum variance that characterizes the data
c. Minimum variance
d. Uniform variance
Answer: b. Maximum variance that characterizes the datagithub - PCA uses which of the following to find principal components?
a. Neural networks
b. Matrix decomposition and eigen decomposition
c. Decision trees
d. Clustering algorithms
Answer: b. Matrix decomposition and eigen decompositiongithub - What happens when you use PCA on a dataset with features measured in very different units without scaling?
a. PCA is unbiased
b. Features with larger scales dominate the principal components
c. No impact on PCA results
d. Features with smaller scales dominate
Answer: b. Features with larger scales dominate the principal componentsgithub - Which component typically explains the least variance?
a. First principal component
b. Last principal component
c. Second principal component
d. Third principal component
Answer: b. Last principal componentgithub - What is the effect of PCA on the correlation matrix of the transformed components?
a. Correlations become zero
b. Correlations increase
c. Correlations become negative
d. Correlation matrix remains the same
Answer: a. Correlations become zerogithub - Which of the following is a common use case of PCA?
a. Image compression
b. Text classification
c. Decision trees
d. Reinforcement learning
Answer: a. Image compressiongeeksforgeeks
- https://www.studocu.com/row/document/tunis-business-school/data-analysis/pca-mcq-student-prepared-mcq/115569708
- https://github.com/Devinterview-io/pca-interview-questions
- https://www.placementpreparation.io/mcq/machine-learning/page/4/
- https://www.geeksforgeeks.org/data-analysis/principal-component-analysis-pca/
- https://www.gatevidyalay.com/tag/principal-component-analysis-questions-and-answers/
- https://www.seas.upenn.edu/~cis520/exams/final2017_solutions.pdf
- https://www.mlstack.cafe/blog/principal-component-analysis-interview-questions
- https://aimcqs.com/dimensionality-reduction
- https://www.scribd.com/document/660120455/ML-Unit-2-mcq
- https://www.sanfoundry.com/1000-machine-learning-questions-answers/
- https://www.codechef.com/learn/course/machine-learning/CLUSTERING/problems/MLPROB202B
- https://iq.opengenus.org/principal-component-analysis-questions/
- https://www.interviewbit.com/machine-learning-mcq/
- https://wayground.com/admin/quiz/65b648a29956be83a45a1d4e/principal-component-analysis-quiz
- https://www.youtube.com/watch?v=GNWFE72cQ6M
- https://www.cs.cmu.edu/~aarti/Class/10701/exams/final2008f-solution.pdf
- https://www.studocu.com/in/document/university-of-mumbai/machine-learning/principal-component-analysis-pca-in-machine-learning/81350092
- https://www.scribd.com/document/919569688/ML-Exam-Q-A