AGIF ORS RM STUDY GROUP 2016, FRANKFURT

摘要

Developed a new dependent demand, network revenue opportunity model (SBQP) which optimizes revenue while providing more stable market-level allocations than existing methods ? In network samples we tested, showed that the SBQP reduces market-level allocation variability over time between i%-8% ? The expected total ROM runtimes increased about 4% using the two-stage approach compared to the SBLP alone ? The reductions in variability were shown to be consistent across a range of tolerances and were strongly statistically significant Future Research ? Conducting customer testing with parallel versions of ROM models to determine if the impacts are noticeable in practice to RM analysts ? Exploring adaptations of the SBQP to other problems such as O&D traffic estimation.