Matching Market

Matching markets analyze the process of pairing agents with preferences, aiming to achieve stable matchings where no pair would prefer to switch partners. Current research focuses on scenarios with incomplete information, where agents' preferences are unknown and must be learned through iterative interactions, often modeled using bandit algorithms like Thompson Sampling or variations of Explore-then-Commit and Upper Confidence Bound. These models are being extended to handle complexities such as two-sided uncertainty, quota constraints, complementary preferences, and time-varying preferences, improving the efficiency and stability of matching in various applications. This work has significant implications for resource allocation, recommendation systems, and online platforms requiring efficient and stable pairing mechanisms.