Optimizing airline seat allocation using reinforcement programming

Low season is always be a big problem for the airlines as the occupant hugely decreased when compared to peak season. Generally, to attract more occupant in low season the airlines will sell more discount fare but the other hand they must balancing the seat allocation to avoid loss. This paper proposed new approach to optimizing the seat allocation. This approach using new algorithm which use the basic concept from reinforcement learning called reinforcement programming. First, we using the previous data to random the initial set of solution. Afterward we enforce some rule and objective function to make initial solution move toward the best solution. Finally, solution will show the revenue and a set of seat allocation for each subclass. We compare result of our idea into several optimization algorithm. The experimental result show the effectiveness of our idea to solve this seat allocation problem.