Monday June 30th 2025, AM & PM
Lecturers: Luca Coraggio, Antonio D’Ambrosio, Maurizio Romano
Introduction to Preference Learning
Topics, AM: Basic preference learning; Introduction to rank data; data visualization; descriptives; distances for rankings. LAB: Use or R packages to handling rank data: data visualization; descriptives; distances. PM: Rank aggregation problem, heuristics, introduction to supervised and unsupervised preference learning. LAB: Rank aggregation problem, examples of supervised and unsupervised preference learning.
Monday June 30th 2025, Evening: Poster Session & Welcome Reception
Tuesday July 1st 2025, AM & PM
Lecturers: Marta Crispino, Anne-Marie George, Øystein Sørensen
Distance-based models for Preference Learning
Topics, AM: The Mallows model, the normalizing constant, ML inference (basic). LAB: The Mallows model, the normalizing constant, ML inference with R and python. Examples from social choice theory. PM: Bayesian inference in the Mallows model. Algorithms. LAB: Bayesian inference in the Mallows model with the BayesMallows package.
Wednesday July 2nd 2025, AM
Lecturers: Cristina Mollica & Luca Tardella
Bayesian Preference Learning via the Plackett-Luce (part I)
Topics: Stagewise models in the ranking literature; A focus on the Plackett-Luce model (PL); Extensions of the basic PL. LAB: The R package PLMIX; Practical examples with R.
Wednesday July 2nd 2025, PM: Social Excursion
Thursday July 3rd 2025, AM
Lecturers: Cristina Mollica & Luca Tardella
Bayesian Preference Learning via the Plackett-Luce (part II)
Topics: Finite PL mixtures: from the MLE to Bayesian inference; Bayesian model comparison criteria; Bayesian model assessment; Handling of label-switching. LAB: The R package PLMIX; Practical examples with R.
Thursday July 3rd 2025, PM
Lecturers: Kate Lee & Geoff Nicholls
Bayesian Inference for Partial Orders from Rank-Order Data (part I)
Topics: What is a Partial Order? Key properties (depth, dimension); Why order partially? Social hierarchies and other applications; Rank-Order data as random linear extensions of a Partial Order; Sequential observation models for Partial Orders. Queue-jumping. Priors for Partial Orders: the bad and the good. LAB: Explore basic Partial Order properties in R.
Thursday July 3rd 2025, Evening: Social Dinner
Friday July 4th 2025, AM
Lecturers: Kate Lee & Geoff Nicholls
Bayesian Inference for Partial Orders from Rank-Order Data (part II)
Topics: The Posterior Distribution. Basic MCMC methods; Quantifying Uncertainty in Partial Order reconstructions; The Royal Acta data and witness lists. Software overview (R code for fitting Partial Orders). LAB: Fit Partial Order models using R. Fixed time and time-series (Python code may also be available for those who prefer/if desired).