Online Rank Elicitation for Plackett-Luce: A Dueling Bandits Approach
Authors: Balázs Szörényi, Róbert Busa-Fekete, Adil Paul, Eyke Hüllermeier
NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In addition to a formal performance and complexity analysis, we present first experimental studies. ... Our approach strongly exploits the assumption of a data generating process that can be modeled by means of a PL distribution. The experimental studies presented in this section are mainly aimed at showing that it is doing so successfully, namely, that it has advantages compared to other approaches in situations where this model assumption is indeed valid. To this end, we work with synthetic data. |
| Researcher Affiliation | Academia | Bal azs Sz or enyi Technion, Haifa, Israel / MTA-SZTE Research Group on Artificial Intelligence, Hungary ... R obert Busa-Fekete, Adil Paul, Eyke H ullermeier Department of Computer Science University of Paderborn Paderborn, Germany |
| Pseudocode | Yes | Algorithm 1 BQS(A, B) ... The pseudocode of the algorithm, called PLPAC, is shown in Algorithm 2. ... Its pseudo-code is shown in Algorithm 3 in Appendix C. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described in this paper. |
| Open Datasets | No | To this end, we work with synthetic data. ... In addition, we conducted some experiments to asses the impact of parameter and to test our algorithms based on Clopper-Pearson confidence intervals. These experiments are deferred to Appendix H and G due to lack of space. |
| Dataset Splits | No | The paper uses synthetic data and does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning into train/validation/test sets. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) 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 like Python 3.8, CPLEX 12.4) needed to replicate the experiment. |
| Experiment Setup | Yes | We tested the learning algorithm by setting the parameters of PL to vi = 1/(c + i) with c = {0, 1, 2, 3, 5}. The parameter c controls the complexity of the rank elicitation task... We evaluated the algorithm on this test case with varying numbers of items M = {5, 10, 15} and with various values of parameter c... (only for c = {0, 2, 5}, the rest of the plots are deferred to Appendix E). ... for M = {5, 10, 15}, δ = 0.05, = 0. |