Temporal Modeling of On-Street Parking Data for Detection of Parking Violation in Smart Cities

Modeling of Parking Violations Using Zero-Inflated Negative Binomial Regression: A Case Study for Berlin

“…This behavior implies a further source of (negative) selection bias. Only a few recent studies have used other data sources like Google Street View ( 12 ), on-street parking sensors ( 13 ), manual field observations ( 14 ), stationary camera setups ( 15 ), or aggregated trajectories of sharing bikes ( 16 ). Various modeling approaches are proposed in the literature to analyze parking violations as for example clustering methods ( 17 ), regression models with spatial dependencies ( 3 ) like spatial lag or spatial error regressions ( 6 ), and multiple linear- ( 12 ) or logit regressions ( 14 ).…”