Rifdatun Ni’mah 

Prodi Sains Data, Fakultas Teknologi Informasi dan Bisnis, Institut Teknologi Telkom Surabaya, Surabaya, Jawa Timur 60231, Indonesia

Regita Putri Permata 

Prodi Sains Data, Fakultas Teknologi Informasi dan Bisnis, Institut Teknologi Telkom Surabaya, Surabaya, Jawa Timur 60231, Indonesia

DOI: https://doi.org/10.19184/mims.v23i2.38203

ABSTRACT 

The study aims to model the distribution of motorcycle departures from the parking lot at peak times with the Poisson process approach. This process involves a discrete number of departures and continuous inter-departure time. The probability distribution candidate was selected to model the data according to the data nature, the stochastic process, and the empirical observation of the departure process. Parameters are estimated using the Maximum Likelihood Estimation (MLE) method and the bootstrapping procedure to construct confidence intervals for the parameter. The goodness-of-fit test is applied to select the best probability distribution that matches empirical data. Inferences to the distribution parameters suggest that Weibull's distribution is more appropriate for describing the motorcycle's inter-departure time. The number of motorcycle departures fits significantly into a negative binomial distribution. The results of the study concluded that the Poisson process applied was a case of overdispersion, with the motorcycle departure rate decreasing over time.

Keywords: Bootstrapping, departure, distribution, goodness-of-fit, Poisson.

MSC2020: 60E05