Exploring Machine Learning Concepts

What is the primary goal of machine learning?

• To enable computers to learn from data and make predictions or decisions.

• To store data without analysis.

• To replace human intelligence entirely.

• To create static algorithms that do not adapt.

Explain the least squares method in linear regression.

• The least squares method in linear regression is a technique used to minimize the sum of the squared differences between observed and predicted values, thereby finding the best-fitting line.

• It focuses on maximizing the differences between observed and predicted values.

• It calculates the average of all observed values.

• It uses the median to find the best-fitting line.

What is the difference between single and multiple variable linear regression?

• Single variable regression is used for classification; multiple variable regression is used for regression tasks.

• Single variable regression analyzes time series data; multiple variable regression analyzes cross-sectional data.

• Single variable regression uses multiple predictors; multiple variable regression uses one predictor.

• Single variable regression uses one predictor; multiple variable regression uses multiple predictors.

Describe the concept of Bayesian linear regression.

• Bayesian linear regression only uses observed data without prior beliefs.

• Bayesian linear regression assumes a fixed set of parameters without any distribution.

• Bayesian linear regression is a method that combines prior beliefs and observed data to estimate the distribution of regression parameters, allowing for uncertainty quantification.

• Bayesian linear regression is a method that does not account for uncertainty in parameter estimates.

How does gradient descent work in optimizing models?

• Gradient descent only works with linear models and cannot optimize complex models.

• Gradient descent optimizes models by iteratively adjusting parameters to minimize the loss function using the negative gradient.

• Gradient descent adjusts parameters based on the positive gradient of the loss function.

• Gradient descent uses random sampling to select parameters for optimization.

What is a discriminant function in linear classification?

• A discriminant function is a graphical representation of data points.

• A discriminant function is a mathematical function used to separate classes in linear classification.

• A discriminant function is a type of neural network architecture.

• A discriminant function is used to calculate the mean of a dataset.

Explain how logistic regression is used for classification tasks.

• Logistic regression requires normally distributed data for accurate predictions.

• Logistic regression is used to predict continuous outcomes.

• Logistic regression is used to model the probability of a binary outcome based on one or more predictor variables.

• Logistic regression can only be applied to datasets with more than two classes.

What is the Naive Bayes classifier and when is it used?

• The Naive Bayes classifier is primarily used for image recognition.

• The Naive Bayes classifier is a clustering algorithm.

• The Naive Bayes classifier is used for regression tasks.

• The Naive Bayes classifier is used for classification tasks, especially in text classification.

How does a support vector machine find the optimal hyperplane?

• By using a linear regression model to fit the data.

• By randomly selecting points in the dataset.

• By minimizing the distance to all data points.

• By maximizing the margin between support vectors.

What are the advantages of using random forests over decision trees?

• Random forests do not provide any insights into feature importance.

• Random forests are slower than decision trees.

• Random forests reduce overfitting, increase accuracy, and provide feature importance.

• Random forests require more data preprocessing than decision trees.