NEW
Font size
WorksheetsRecap Machine Learning
Total questions: 8
Worksheet time: 4mins
Which of the following statements is true for supervised learning?
The algorithm learns from input-output pairs to predict outputs for unseen inputs.
Supervised learning does not require any labeled data.
Supervised learning can only be used for classification tasks.
In supervised learning, the output space must be continuous.
Which of the following is a key characteristic of a probabilistic classifier?
It predicts the class label based solely on distance metrics.
It provides a probability distribution over class labels for each prediction.
It uses a decision boundary to separate classes in the feature space.
It works by minimizing the error between predicted and true values.
Which of the following techniques is commonly used for hyperparameter tuning in machine learning models to improve performance?
Grid Search
Cross-validation
Data augmentation
Regularization
Which of the following metrics would be most appropriate for evaluating the performance of a regression model?
Accuracy
Precision
Mean Absolute Error (MAE)
F1-score
Which of the following deep learning architectures is primarily used for image classification tasks due to its ability to capture spatial hierarchies in images?
Recurrent Neural Networks (RNN)
Fully Connected Neural Networks (FCNN)
Support Vector Machines (SVM)
Convolutional Neural Networks (CNN)
What does generalization refer to in machine learning?
The model's ability to memorize training data
The model's ability to perform well on new, unseen data
The process of increasing training accuracy
Using more features to improve performance
In machine learning, overfitting means:
The model works well on every data point
The model is too simple to learn patterns
The model learns training data too exactly
The model always gives the right results
Which of the following is true about a neural network?
It follows only fixed logic rules to decide
It is modeled after the human brain system
It does not require any input data to trainIt works without any training data needed
It is useful only for working with pictures
