Filtering Infrequent Behavior in Business Process Discovery by Using the Minimum Expectation

2. Related Work

A number of outlier detection algorithms have been proposed in the data mining field. These algorithms build a data model (e.g., a statistical, linear, or probabilistic model) that describes the normal behavior and considers all data points that deviate from this model as outliers (Aggarwal et al., 2015).

In the context of temporal data, these algorithms have been extensively surveyed by Gupta et al. (2014) (for events with continuous values, known as time series) and by Chandola et al. (2012) (for events with discrete values, known as discrete sequences).