PREDICTIVE MODELS FOR WORK ORDERS IN IFES: MACHINE LEARNING STRATEGIES FOR RESOURCE MANAGEMENT

Advances in public transport are important for urban mobility, as they are essential to indicate the level of development of a region. This article proposes a flow-based mathematical programming model for the bus crew allocation problem that is based on six operational stages: (i) interviewing urban transport companies; (ii) generate and implement a model for the service scale problem; (iii) select instances for testing; (vi) evaluate and adjust the model; (v) validate the mathematical programming model; and (vi) implement Fix-and-Optimize heuristics without an exact model and compare the results. This work contributes to the progress of the area and motivates the development of the urban transportation sector. Only computational results are presented, based on real-life instances of a Brazilian bus company, and a flux-based model for the Crew Scheduling Problem (CSP).