Q.1 What is the main characteristic of supervised learning?
It uses unlabeled data
It uses labeled data
It does not require training
It only clusters data
Explanation - Supervised learning relies on labeled data, where each input has a corresponding correct output used to train models.
Correct answer is: It uses labeled data
Q.2 Which of the following is a supervised learning algorithm?
K-means
Linear Regression
Apriori
DBSCAN
Explanation - Linear Regression is a supervised learning algorithm that maps inputs to continuous outputs.
Correct answer is: Linear Regression
Q.3 In supervised learning, the dataset consists of:
Only features
Only labels
Features and corresponding labels
Random data
Explanation - Supervised learning requires both features (inputs) and labels (outputs) for training.
Correct answer is: Features and corresponding labels
Q.4 Which type of supervised learning predicts continuous values?
Classification
Regression
Clustering
Association
Explanation - Regression models predict continuous values like house prices or stock values.
Correct answer is: Regression
Q.5 Which type of supervised learning predicts discrete labels?
Regression
Classification
Clustering
Dimensionality Reduction
Explanation - Classification assigns inputs into discrete categories such as spam or not spam.
Correct answer is: Classification
Q.6 What does overfitting mean in supervised learning?
Model performs poorly on training data
Model fits noise and fails on test data
Model is too simple
Model ignores features
Explanation - Overfitting occurs when the model memorizes training data and fails to generalize.
Correct answer is: Model fits noise and fails on test data
Q.7 Which metric is commonly used to evaluate classification models?
Mean Squared Error
Accuracy
Silhouette Score
Lift
Explanation - Accuracy measures the proportion of correctly predicted labels in classification problems.
Correct answer is: Accuracy
Q.8 Which metric is best for regression problems?
Accuracy
Mean Squared Error
F1 Score
Precision
Explanation - MSE measures average squared difference between predicted and actual values in regression.
Correct answer is: Mean Squared Error
Q.9 Which algorithm is commonly used for classification tasks?
KNN
Apriori
PCA
K-means
Explanation - K-Nearest Neighbors (KNN) is often used for classification problems.
Correct answer is: KNN
Q.10 What does the 'supervised' in supervised learning refer to?
Supervision by human during training
Using labeled examples
Using unsupervised features
Using random selection
Explanation - The term 'supervised' comes from having correct outputs (labels) provided during training.
Correct answer is: Using labeled examples
Q.11 Decision Trees are an example of which learning type?
Supervised Learning
Unsupervised Learning
Reinforcement Learning
Evolutionary Learning
Explanation - Decision Trees use labeled data, making them supervised learning algorithms.
Correct answer is: Supervised Learning
Q.12 In classification, labels are usually:
Continuous values
Categories
Random values
Unsupervised clusters
Explanation - Classification assigns data into discrete categories like 'yes/no'.
Correct answer is: Categories
Q.13 Which of the following is NOT supervised learning?
Logistic Regression
Linear Regression
K-means
Support Vector Machines
Explanation - K-means is an unsupervised clustering algorithm.
Correct answer is: K-means
Q.14 Support Vector Machines (SVM) are typically used for:
Classification
Clustering
Dimensionality Reduction
Association Rule Mining
Explanation - SVM is a supervised algorithm mainly used for classification tasks.
Correct answer is: Classification
Q.15 What is the role of the test set in supervised learning?
Used to train the model
Used to tune hyperparameters
Used to evaluate performance
Used to store data
Explanation - The test set evaluates how well the model generalizes to unseen data.
Correct answer is: Used to evaluate performance
Q.16 Which of these is an example of regression?
Predicting house prices
Classifying emails as spam
Grouping students
Finding frequent patterns
Explanation - House prices are continuous values, suitable for regression.
Correct answer is: Predicting house prices
Q.17 Which algorithm is suitable for predicting customer churn (yes/no)?
Linear Regression
Logistic Regression
K-means
PCA
Explanation - Logistic Regression predicts binary outcomes, such as churn or no churn.
Correct answer is: Logistic Regression
Q.18 What is the key assumption in supervised learning?
Data is unlabeled
Labels are known
Noisy data is required
Data is clustered
Explanation - Supervised learning assumes each input has a corresponding known label.
Correct answer is: Labels are known
Q.19 What happens if a model underfits?
It memorizes the data
It fails to capture patterns
It performs well on unseen data
It has high variance
Explanation - Underfitting occurs when the model is too simple and fails to learn the underlying patterns.
Correct answer is: It fails to capture patterns
Q.20 Which algorithm works by finding a hyperplane to separate classes?
Naive Bayes
SVM
KNN
Random Forest
Explanation - Support Vector Machine separates data using an optimal hyperplane.
Correct answer is: SVM
Q.21 Which algorithm is based on probability in supervised learning?
Naive Bayes
K-means
Apriori
PCA
Explanation - Naive Bayes uses Bayes’ theorem to classify data based on probabilities.
Correct answer is: Naive Bayes
Q.22 What is the main difference between supervised and unsupervised learning?
Supervised uses labels, unsupervised does not
Unsupervised uses labels, supervised does not
Both use labels
Neither use labels
Explanation - Supervised learning requires labels, while unsupervised learning works without labels.
Correct answer is: Supervised uses labels, unsupervised does not
Q.23 Which of these is an ensemble supervised learning method?
Random Forest
K-means
Apriori
DBSCAN
Explanation - Random Forest is an ensemble method built on multiple decision trees.
Correct answer is: Random Forest
Q.24 Which dataset split ratio is commonly used in supervised learning?
90/10
70/30
50/50
10/90
Explanation - A common practice is to use 70% of data for training and 30% for testing.
Correct answer is: 70/30
Q.25 What type of problem is predicting exam grades (A, B, C, D)?
Classification
Regression
Clustering
Association
Explanation - Grades are discrete categories, so it is a classification problem.
Correct answer is: Classification