Machine Learning & Predictive Analytics Flashcards

Which of the following is an unsupervised learning task?

Answer: Grouping customers into segments without labels

Clustering customers without predefined labels is a classic unsupervised learning task.

What does regularization (L1 or L2) primarily do in a regression model?

Answer: Penalizes large coefficients to reduce overfitting

Regularization adds a penalty on coefficient size, discouraging overly complex models and reducing overfitting.

L1 regularization (Lasso) is especially useful because it can do what?

Answer: Drive some coefficients exactly to zero for feature selection

Lasso can shrink some coefficients to exactly zero, effectively performing automatic feature selection.

In k-means clustering, how is the number of clusters typically chosen?

Answer: Using the elbow method on within-cluster variance

The elbow method plots within-cluster variance against k and picks the point of diminishing returns.

What problem does principal component analysis (PCA) primarily address?

Answer: Dimensionality reduction

PCA reduces the number of features by projecting data onto directions of maximum variance.

A model shows high error on both training and test sets. What is the most likely issue?

Answer: Underfitting

High error on both sets indicates the model is too simple to capture the underlying pattern, i.e., underfitting.

Which approach best prevents data leakage when scaling features?

Answer: Fit the scaler only on training data, then apply to test data

Fitting the scaler only on training data prevents test-set information from leaking into the model.