New ORBIT method optimizes dynamic pricing with semiparametric models

Researchers have developed a new method called ORBIT for contextual dynamic pricing in semiparametric models. This approach leverages the structure of an oracle price map, which is learned through bandit convex optimization within a trust region. The method aims to minimize regret, achieving a bound of approximately O(T^((2β-1)/(4β-3)) + sqrt(dT)). ORBIT can be extended to handle sparse high-dimensional linear utilities and nonparametric Hölder utilities. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel optimization technique for dynamic pricing that could improve efficiency in e-commerce and other applications.

RANK_REASON Academic paper detailing a new method for contextual dynamic pricing. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• arXiv
• ORBIT

• paper
• other