Click the "Show Details" button in each presentation to view detailed information including abstracts, authors, affiliations, and keywords.
Keynote 2 1 presentations
The clustered Mallows model
Abstract:
Rank data often arises where assessors rank items in decreasing order of preference. However, strict preference relations can be unrealistic for real data. In large item sets, assessors might prioritise certain items, rank others low, and express indifference towards the remaining. Rank aggregation may involve decisive preferences between some items and ambiguity in others. In this talk, we extend the famous Mallows model (Mallows, Biometrika 1957) to accommodate item indifference. The Clustered Mallows model, which we propose allows one to learn ranked clusters of items which items within a cluster each share the same tied rank. We will outline some statistical challenges associated with this model as well an illustration of its performance in practice.
Rank data often arises where assessors rank items in decreasing order of preference. However, strict preference relations can be unrealistic for real data. In large item sets, assessors might prioritise certain items, rank others low, and express indifference towards the remaining. Rank aggregation may involve decisive preferences between some items and ambiguity in others. In this talk, we extend the famous Mallows model (Mallows, Biometrika 1957) to accommodate item indifference. The Clustered Mallows model, which we propose allows one to learn ranked clusters of items which items within a cluster each share the same tied rank. We will outline some statistical challenges associated with this model as well an illustration of its performance in practice.