The case arrival model is implemented by inferring a probability distribution to model the inter-arrival times. First, all inter-arrival times (duration from one case start to the next one) are computed, then SIMOD filters outliers and infers a probability distribution.
Some processes benefit from a stricter outlier filtering, others from a more relaxed one. Set the filtering threshold as a parameter to optimize in the HyperOpt process. Consider just fixing 4-5 values by default (e.g., 5.0, 10.0, 20.0, 50.0, 100.0).
The case arrival model is implemented by inferring a probability distribution to model the inter-arrival times. First, all inter-arrival times (duration from one case start to the next one) are computed, then SIMOD filters outliers and infers a probability distribution.
Some processes benefit from a stricter outlier filtering, others from a more relaxed one. Set the filtering threshold as a parameter to optimize in the HyperOpt process. Consider just fixing 4-5 values by default (e.g., 5.0, 10.0, 20.0, 50.0, 100.0).