“Adaptive Learning to Prevent Spiral-Down in Branded-Fare Revenue Management”

Department of Analytics and Operations

A short report from Wu Lan on her paper in the 2025 INFORMS conference.

I attended the 2025 INFORMS Annual Meeting held from October 26 – 29, 2025 at Georgia World Congress Center and Omni Atlanta Hotel at Centennial Park, USA and presented my paper in a dedicated session and received targeted feedback from industry practitioners in revenue management.

I attended several high-quality sessions, notably the Nicholson Student Paper Award presentations, which showcased innovative modeling and data-driven methods. These talks were especially useful for sharpening my presentation techniques (framing the problem, visualizing results, and pacing Q&A;) and for sensing current research trends—e.g., robust learning for operations, interpretable machine learning in RM, and experimentation at scale.

Beyond sessions, I connected with PhD students from other universities, exchanging ideas on methodology, datasets, and the job market process. These conversations helped identify potential collaborators and venues for seminar exchanges.

Key takeaways and next steps: Incorporate industry feedback on deployment metrics and system integration; tighten the narrative and figures to align with best practices seen in award sessions; and follow up with new contacts to explore collaboration and cross-institution seminar