Parameter Estimation for Generalized Thurstone Choice Models

Authors: Milan Vojnovic, Seyoung Yun

ICML 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Figure 1. Mean square error for two different generalized Thurstone choice models TF : (left) F is a double-exponential distribution, and (right) F is a uniform distribution. The vertical bars denote 95% confidence intervals. The results confirm two qualitatively different relations with the cardinality of comparison sets as suggested by the theory.
Researcher Affiliation Industry Milan Vojnovic MILANV@MICROSOFT.COM Microsoft Research, Cambridge, UK
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any concrete access to source code for the methodology described.
Open Datasets No The paper describes generating synthetic comparison sets by 'independent uniform random samples from the set of all items' for its numerical examples, but it does not use or provide concrete access information for a publicly available dataset.
Dataset Splits No The paper describes a simulation experiment but does not provide specific details on training, validation, or test dataset splits. It describes generation of synthetic data.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment.
Experiment Setup Yes We fix the values of the number of items n and the number of comparisons m, and consider a choice of a generalized Thurstone model TF for the value of parameter = 0. We consider comparison sets of the same cardinality of value k that are independent uniform random samples from the set of all items. For every fixed value of k, we run 100 repetitions to estimate the mean square error. We do this for the distribution of noise according to a double-exponential distribution (Bradley-Terry model) and according to a uniform distribution, both with unit variance.