Research And Development Intern
Current- Formulate a decision recommendation framework for Integrated Airline Disruption Recovery problem, which attempts to minimize the operational cost in case of disruptions through readjusting and rescheduling the involved entities, including flights, aircraft, crew and passengers, in a systematic and informed manner.- Establish a novel solution framework by incorporating the state-of-the-art Deep Reinforcement Learning (DRL) into the standard heuristic approach, Variable Neighborhood Search (VNS), which features well-designed neighborhood structures and state evaluator.- Experimental results show that the objective value generated by our approach is within a 1.5% gap with respect to the optimal objective of the CPLEX solver, with significant improvement regarding runtime.
