Machine Learning Quiz Questions and Answers (original) (raw)

What is overfitting in the context of machine learning models?

• Fitting a model with insufficient data
• Fitting a model too closely to the training data
• Fitting a model with too few features
• Fitting a model to the validation set

In reinforcement learning, what is the role of the exploration-exploitation trade-off?

• Balancing the use of supervised and unsupervised learning
• Balancing the trade-off between precision and recall
• Balancing the trade-off between exploring new actions and exploiting known actions

How does the choice of a loss function impact the training of a machine learning model?

• The loss function has no impact on training
• The loss function determines the optimization objective
• The loss function defines the model's architecture
• The choice of loss function only impacts model evaluation

Explain the concept of "latent variables" in probabilistic graphical models.

• Variables that are not observed directly but inferred from observed variables
• Variables that are directly measured in the dataset
• Variables representing missing data
• Variables used to encode temporal information

What is the difference between bagging and boosting in ensemble learning?

• Bagging increases model diversity, boosting decreases it
• Bagging trains models sequentially, boosting trains them in parallel
• Bagging combines predictions using voting, boosting combines predictions using weighted averaging
• Bagging trains each model independently, boosting focuses on examples misclassified by previous models

What is the concept of entropy in the context of decision trees?

• The measure of impurity or disorder in a set of data
• The depth of the decision tree
• The ratio of training to testing data
• The number of leaf nodes in the tree

What is the purpose of the Expectation-Maximization (EM) algorithm in unsupervised learning?

• Maximizing the likelihood of the observed data
• Minimizing the reconstruction error in autoencoders
• Imputing missing values in a dataset
• Iteratively estimating parameters for mixture models

What is the role of the learning_rate parameter in gradient descent optimization?

• The speed at which the algorithm converges
• The regularization strength applied to the model
• The number of iterations in the optimization process
• The size of the steps taken during each iteration

What is the purpose of the epochs parameter in neural network training?

• The number of layers in the neural network
• The number of training examples processed in one iteration
• The learning rate for weight updates
• The number of complete passes through the entire training dataset

How does the choice of a kernel impact the performance of a Support Vector Machine (SVM)?

• The kernel has no impact on SVM performance
• The kernel determines the SVM's maximum margin
• The kernel defines the transformation of input features into a higher-dimensional space
• The kernel influences the learning rate during training

There are 32 questions to complete.

Take a part in the ongoing discussion