Unsupervised Learning Flashcards

What is the role of the 'epsilon' parameter in DBSCAN?

Answer: It defines the maximum radius of a neighborhood around a point

Epsilon (ε) defines the radius within which DBSCAN searches for neighboring points to determine core points and cluster membership.

Which technique can reduce dimensionality while also revealing nonlinear structure in data?

Answer: UMAP

UMAP (Uniform Manifold Approximation and Projection) is a nonlinear dimensionality reduction technique that preserves both local and some global structure.

What is a key difference between agglomerative and divisive hierarchical clustering?

Answer: Agglomerative starts with each point as its own cluster and merges; divisive starts with one cluster and splits

Agglomerative (bottom-up) clustering merges the closest clusters iteratively, while divisive (top-down) starts with all data in one cluster and recursively splits.

In the context of anomaly detection, which unsupervised approach assigns anomaly scores based on how easily a point can be isolated?

Answer: Isolation Forest

Isolation Forest isolates observations by randomly partitioning data; anomalies are isolated in fewer steps (shorter path lengths) than normal points.

What is the 'reconstruction error' in an autoencoder used as a signal for?

Answer: Detecting anomalies, since unusual inputs have higher reconstruction error

Autoencoders trained on normal data learn to reconstruct normal inputs well; anomalous inputs result in higher reconstruction error, flagging them as outliers.

Which of the following best describes 'feature learning' in unsupervised learning?

Answer: Automatically discovering useful representations from raw data without labels

Feature learning (representation learning) involves algorithms like autoencoders and deep belief networks that discover compact, meaningful representations from unlabeled data.

What is a Variational Autoencoder (VAE) primarily used for compared to a standard autoencoder?

Answer: Generating new samples by learning a smooth latent space distribution

VAEs impose a probabilistic distribution (typically Gaussian) on the latent space, enabling generation of new, realistic samples by sampling from that distribution.