R. E. Lillo, P. Ramirez Cobo, M. González Bernal
Markovian arrival process (MAPs) are known to constitute a versatile class of point processes that allow for dependent inter-arrival times. In this work, we aim to exploit such property for modeling modern call centers, which are characterized by non-negligible dependence patterns. Most of previous statistical approaches for MAPs are based on the distribution of the inter-arrival times. In this work, however, a different perspective is adopted, being the inference focused on the properties of the associated counting process. In particular, this work deepens into the covariance function of the counting process for which almost the only thing that is known is its closed-form
expression. New properties concerning the correlation patterns and monotonicity shall be illustrated.
Keywords: MAP Process, Counting Process, Call center, Inference in Point Process
Scheduled
GT17 Stochastic Processes and their Applications II
June 7, 2022 3:30 PM
A26
