Anomaly Detection Quiz

What is anomaly detection primarily used for?

• Identifying common patterns in data

• Clustering similar data points together

• Identifying patterns or instances in data that deviate significantly from the norm

• Reducing data dimensionality

Which of the following is NOT a typical application of anomaly detection?

• Fraud detection

• Market basket analysis

• Intrusion detection

• System health monitoring

Anomalies in data are often referred to as:

• Clusters

• Noise

• Outliers

• Labels

Which of the following describes a scenario that could be identified by anomaly detection?

• Grouping customers based on purchase history

• Detecting a transaction of $10,000 from a customer who usually spends $100 per transaction

• Predicting stock prices

• Classifying emails as spam or not

What type of data is typically used for novelty detection?

• Data containing both normal and anomalous points

• Data with missing values

• Data presumed to be "clean"

• Data with mixed types

In a Gaussian Mixture Model (GMM), what does each Gaussian component represent?

• A cluster of data points

• A single data point

• An outlier

• A decision boundary

How does a GMM differ from K-Means clustering?

• GMMs assume clusters are spherical and equally sized

• GMMs are non-probabilistic

• GMMs can model elliptical clusters with different shapes and densities

• GMMs are faster and more scalable

In the context of anomaly detection using GMMs, a data point with very low likelihood (density) is likely to be considered:

When using anomaly detection in manufacturing, a critical density threshold is set based on:

• The size of the dataset

• The proportion of expected defective items

• The number of features in the dataset

• The type of sensors used

What is the primary difference between supervised and unsupervised anomaly detection?

• Supervised methods require labeled data, while unsupervised methods do not

• Supervised methods are faster than unsupervised methods

• Unsupervised methods are only used for large datasets

• Unsupervised methods require labeled data, while supervised methods do not