D491 Introduction to Analytics - Set 4 - Part 1

Correct Answer: Recursive feature elimination

Recursive feature elimination is a feature selection method that recursively removes the least important features to build a model with fewer inputs.

Correct Answer: To model complex, non-linear relationships in data

Neural networks are capable of modeling complex and non-linear relationships, making them highly useful for tasks such as image recognition and language processing.

Correct Answer: Histogram

Histograms are used to visualize the distribution of a single continuous variable by showing the frequency of data points in specified ranges.

Correct Answer: Maximizing a reward through trial and error

Reinforcement learning involves an agent interacting with an environment and learning the best actions to take by maximizing rewards over time.

Correct Answer: F1 Score

The F1 score is the harmonic mean of precision and recall, making it a useful metric for imbalanced classification tasks where accuracy might be misleading.

Correct Answer: To find the best hyperparameters for a model

Grid search is a technique used to find the optimal combination of hyperparameters for a machine learning model by evaluating different parameter combinations.

Correct Answer: Textual data

NLP techniques are used to process, analyze, and derive meaningful information from textual data such as documents, tweets, or conversations.

Correct Answer: The separation between clusters

The silhouette coefficient measures how well an object is clustered by comparing its proximity to other points within its own cluster versus points in other clusters.

Correct Answer: k-Nearest Neighbors (k-NN)

k-NN is widely used in recommendation systems for collaborative filtering by recommending items based on the preferences of similar users.

Correct Answer: To reduce variance

Bagging reduces variance by training multiple models on different subsets of the data and combining their predictions, leading to a more stable model.

Correct Answer: Principal Component Analysis (PCA)

PCA is a dimensionality reduction technique that transforms data into a smaller set of uncorrelated variables called principal components.

Correct Answer: Support Vector Machine (SVM)

SVMs work by finding the optimal hyperplane that best separates data points of different classes in a high-dimensional space.

Correct Answer: To prevent overfitting by penalizing large coefficients

L2 regularization (also known as Ridge) adds a penalty to the size of the coefficients, discouraging overly complex models and helping to prevent overfitting.

Correct Answer: Logistic regression

Logistic regression is used for binary classification tasks where the outcome variable has two possible categories.

Correct Answer: To improve the performance of machine learning models

Feature engineering involves creating new features from existing data to enhance the predictive performance of machine learning models.

Correct Answer: Binary encoding

Binary encoding is a technique used to reduce the number of dimensions introduced by high cardinality categorical variables by converting them into binary digits.

Correct Answer: Mean squared error (MSE)

MSE measures the average squared difference between the actual and predicted values, making it a common metric for evaluating regression models.

Correct Answer: It arranges data into a tree-like structure

Hierarchical clustering creates a tree-like structure (dendrogram) where data points are grouped into nested clusters.

Correct Answer: It reduces overfitting

Random forests reduce overfitting by averaging the predictions of multiple decision trees trained on different subsets of the data.

Correct Answer: Singular Value Decomposition (SVD)

SVD is a technique used for reducing the dimensionality of a dataset by decomposing a matrix into its singular vectors and values.