Anomaly Detection

Question 1

Anomaly Detection / Deviation Detection / Exception Mining

Answer 1

The process of identifying patterns or instances in data that deviate significantly from the norm or expected behaviour.

Question 2

2 Characteristics of Anomaly Detection Methods

Answer 2

Model-based - Learns a model of both the normal and anomalous classes.
Model-free

Question 3

2 Type of Model-Based Anomaly Detection

Answer 3

Unsupervised - Anomalies are those points that don’t fit well
Supervised - Anomalies are regarded as a rare class

Question 4

2 Anomaly Detection Techniques

Answer 4

Statistical Approaches
Gaussian Mixture Model (GMM)
The instances located in areas below that threshold density => outliers

Proximity-based
Distance-Based Approaches - The outlier score of an object is the distance to its kth nearest neighbour
Proximity-based approaches - Local perspective

Question 5

Anomaly Detection vs Novelty Detection

Answer 5

Anomaly Detection - The algorithm is trained on a dataset that may contain outliers, and the goal is typically to identify these outliers (within the training set), as well as outliers among new instances.

Novelty Detection - The algorithm is trained on a dataset that is presumed to be “clean,” and the objective is to detect novelties strictly among new instances.