Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: On the identifiability of mixtures of ranking models
(Submitted on 31 Jan 2022 (v1), last revised 24 May 2022 (this version, v2))
Abstract: Mixtures of ranking models are standard tools for ranking problems. However, even the fundamental question of parameter identifiability is not fully understood: the identifiability of a mixture model with two Bradley-Terry-Luce (BTL) components has remained open. In this work, we show that popular mixtures of ranking models with two components (BTL, multinomial logistic models with slates of size 3, or Plackett-Luce) are generically identifiable, i.e., the ground-truth parameters can be identified except when they are from a pathological subset of measure zero. We provide a framework for verifying the number of solutions in a general family of polynomial systems using algebraic geometry, and apply it to these mixtures of ranking models to establish generic identifiability. The framework can be applied more broadly to other learning models and may be of independent interest.
Submission history
From: Xucheng Zhang [view email][v1] Mon, 31 Jan 2022 11:23:31 GMT (48kb)
[v2] Tue, 24 May 2022 15:16:22 GMT (56kb)
Link back to: arXiv, form interface, contact.