WorksheetsML OVERVIEW
Total questions: 65
Worksheet time: 37mins
What is the primary goal of machine learning?
Write explicit programs for tasks
Automate analytical model building
Store large amounts of data
Perform arithmetic calculations
Which of the following is NOT a type of machine learning?
Supervised Learning
Unsupervised Learning
Reinforcement Learning
Deterministic Learning
Which of the following is an example of supervised learning?
Clustering customers by behavior
Classifying emails as spam or not spam
Reducing image dimensions
Detecting network anomalies
What type of learning involves discovering hidden patterns in data without labels?
Supervised Learning
Semi-supervised Learning
Unsupervised Learning
Reinforcement Learning
Which application is NOT a typical use case of machine learning?
Fraud detection
Prediction of equipment failures
Generating prime numbers
Recommendation engines
What are the two main types of supervised learning?
Regression and Classification
Clustering and Dimensionality Reduction
Regression and Clustering
Anomaly Detection and Classification
What algorithm is commonly used for classification tasks?
Principal Component Analysis
Linear Regression
Logistic Regression
K-Means Clustering
In supervised learning, the dataset contains:
Only input features
Only output labels
Input features and corresponding output labels
No labeled data
Which of the following is an example of a regression problem?
Predicting house prices
Detecting spam emails
Grouping customers by behavior
Recommending movies
Which is NOT an example of supervised learning?
Predicting stock prices
Fraud detection
Customer segmentation
Spam email detection
Which technique is used for dimensionality reduction in unsupervised learning?
Logistic Regression
Principal Component Analysis
Random Forest
Gradient Descent
What is the primary goal of clustering in unsupervised learning?
Predicting output values
Reducing features in a dataset
Grouping similar data points
Detecting anomalies
Which algorithm is used for clustering tasks?
K-Means
Decision Trees
Logistic Regression
Random Forest
Which of the following is NOT an unsupervised learning task?
Clustering
Dimensionality Reduction
Anomaly Detection
Predicting house prices
What type of unsupervised learning involves detecting outliers?
Dimensionality Reduction
Clustering
Anomaly Detection
Classification
What happens when a model overfits?
It captures the general pattern but not the details
It performs well on test data but poorly on training data
It learns noise and performs poorly on new data
It is too simple to capture patterns in the data
Underfitting occurs when:
The model is too complex
The model captures noise instead of patterns
The model is too simple to learn patterns
There are too many features in the dataset
Which of the following is a sign of overfitting?
High training accuracy and high test accuracy
Low training accuracy and high test accuracy
High training accuracy and low test accuracy
Low training accuracy and low test accuracy
What can help reduce underfitting?
Simplifying the model
Increasing the training time
Reducing the number of features
Adding more data
Which is NOT a common cause of overfitting?
Too many features
Training the model for too many iterations
Insufficient training data
Using simpler models
The confusion matrix is used to:
Evaluate regression models
Measure the variance of predictions
Compare actual and predicted values in classification
Select hyperparameters
What does the True Positive (TP) in a confusion matrix represent?
Incorrectly predicted positives
Correctly predicted negatives
Correctly predicted positives
Incorrectly predicted negatives
Precision measures:
How many actual positives were correctly identified
How many predicted positives are actually positive
The proportion of errors in the predictions
The ability of the model to generalize
Which of the following metrics balances precision and recall?
Accuracy
F1 Score
Recall
Precision
Mean Squared Error (MSE) gives more weight to:
Smaller errors
Larger errors
Average errors
No errors
What is the primary advantage of Root Mean Squared Error (RMSE)?
It treats all errors equally
It has the same scale as the target variable
It is unaffected by outliers
It is easier to calculate than MSE
R-squared (R²) measures:
The variance in independent variables
The variance explained by the model
The total error in predictions
The average of absolute errors
Which metric is sensitive to outliers?
Mean Absolute Error (MAE)
Mean Squared Error (MSE)
R-squared (R²)
Accuracy
What does an R² score of 0 indicate?
The model perfectly predicts the data
The model performs worse than the mean
The model performs no better than the mean
The model performs perfectly on training data
Which application is an example of clustering?
Predicting stock prices
Grouping customers based on purchase behavior
Classifying emails as spam
Detecting fraudulent transactions
Which task involves anomaly detection?
Customer segmentation
Fraud detection
Predicting house prices
Sentiment analysis
Which is an example of supervised learning?
Predicting credit scores
Segmenting customer groups
Reducing image size
Detecting anomalies
Text sentiment analysis is an example of:
Clustering
Regression
Classification
Dimensionality reduction
Overfitting occurs when a model learns the ___ and ___ in the training data, resulting in poor generalization to new data.
(a)
Unsupervised learning aims to identify ___ and ___ in data without the use of labeled outputs.
(a)
In supervised learning, ___ algorithms predict continuous values, while ___ algorithms predict discrete categories.
(a)
The (a) is the harmonic mean of precision and recall.
K-Means, DBSCAN, and hierarchical clustering are common algorithms used for (a) in unsupervised learning.
Mean Absolute Error (MAE) measures the average of the (a) differences between predicted and actual values.
Dimensionality reduction is often used for ___ and ___ in machine learning.
(a)
An R² score of (a) indicates that the model perfectly fits the data.
Supervised learning requires labeled data for training.
True
False
A high accuracy score always indicates a good model.
true
false
Overfitting occurs when a model is too simple for the given data.
True
False
Recall measures the proportion of actual positives correctly identified.
True
False
PCA is an algorithm used for clustering in unsupervised learning.
true
false
The Mean Squared Error (MSE) gives more weight to larger errors compared to MAE.
True
False
What is the main characteristic of supervised learning?
No labeled data is used
Both input and corresponding outputs are used
Only input data is provided
Patterns are detected automatically
In supervised learning, which of the following algorithms is used for regression tasks?
Logistic Regression
Decision Trees
K-Means Clustering
PCA
Which of the following is a primary task of unsupervised learning?
Classification
Clustering
Regression
Prediction
What kind of data is used in unsupervised learning?
Data with labels
Data without labels
Data with missing values
Data with continuous outputs
A model achieves 99% accuracy on training data but only 50% accuracy on test data. What is the issue?
Underfitting
Overfitting
Poor Data Cleaning
High Bias
What is the formula for Recall?
TP / (TP + FP)
TP / (TP + FN)
FP / (FP + TP)
TN / (TN + FP)
What is the purpose of splitting data into training and test sets?
To avoid data leakage
To optimize model parameters
To evaluate the model's performance on unseen data
To improve training accuracy
What is the feedback loop in the machine learning process used for?
To improve test set performance
To tweak the model and re-train if the performance is poor
To deploy the model
To validate the final model
Which unsupervised learning algorithm would you use to group customers based on purchasing behavior?
Decision Trees
K-Means
Logistic Regression
PCA
What is the main purpose of dimensionality reduction?
Predict categorical values
Improve training speed and visualization
Increase the number of features
Reduce overfitting
Which of the following is NOT an unsupervised learning application?
Customer segmentation
Fraud detection
Anomaly detection
Dimensionality reduction
Which regression metric gives higher weight to larger errors?
MAE
MSE
Accuracy
R-Squared
A regression model has an R² score of -0.5. What does this indicate?
The model is overfitting
The model is better than the mean
The model performs worse than predicting the mean
The model is perfect
What is the first step in the Machine Learning Process?
Data Cleaning
Data Acquisition
Model Deployment
Model Testing
Which step directly follows "Model Training & Building"?
Data Acquisition
Model Deployment
Model Testing
Data Cleaning
What is the purpose of the Test Data step?
To train the model
To evaluate the model’s generalization performance
To clean the raw data
To deploy the model into production
In which step does the model interact with real-world data?
Model Testing
Model Training & Building
Data Cleaning
Model Deployment
What is the role of "Data Cleaning" in the Machine Learning Process?
It evaluates the model’s accuracy
It removes noise and prepares data for training
It splits the data into training and testing sets
It adjusts model hyperparameters
