Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
120 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Spectral Ranking Inferences based on General Multiway Comparisons (2308.02918v3)

Published 5 Aug 2023 in stat.ME, cs.IT, cs.LG, math.IT, math.ST, stat.ML, and stat.TH

Abstract: This paper studies the performance of the spectral method in the estimation and uncertainty quantification of the unobserved preference scores of compared entities in a general and more realistic setup. Specifically, the comparison graph consists of hyper-edges of possible heterogeneous sizes, and the number of comparisons can be as low as one for a given hyper-edge. Such a setting is pervasive in real applications, circumventing the need to specify the graph randomness and the restrictive homogeneous sampling assumption imposed in the commonly used Bradley-Terry-Luce (BTL) or Plackett-Luce (PL) models. Furthermore, in scenarios where the BTL or PL models are appropriate, we unravel the relationship between the spectral estimator and the Maximum Likelihood Estimator (MLE). We discover that a two-step spectral method, where we apply the optimal weighting estimated from the equal weighting vanilla spectral method, can achieve the same asymptotic efficiency as the MLE. Given the asymptotic distributions of the estimated preference scores, we also introduce a comprehensive framework to carry out both one-sample and two-sample ranking inferences, applicable to both fixed and random graph settings. It is noteworthy that this is the first time effective two-sample rank testing methods have been proposed. Finally, we substantiate our findings via comprehensive numerical simulations and subsequently apply our developed methodologies to perform statistical inferences for statistical journals and movie rankings.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (55)
  1. The approximability of assortment optimization under ranking preferences. Operations Research, 66 1661–1669.
  2. A revealed preference ranking of us colleges and universities. The Quarterly Journal of Economics, 128 425–467.
  3. Generalized method-of-moments for rank aggregation. Advances in Neural Information Processing Systems, 26.
  4. Group recommendations with rank aggregation and collaborative filtering. In Proceedings of the fourth ACM conference on Recommender systems.
  5. The Netflix Prize. In Proceedings of KDD cup and workshop, vol. 2007. New York.
  6. Bayesian nonparametric Plackett–Luce models for the analysis of preferences for college degree programmes. The Annals of Applied Statistics, 8 1145–1181.
  7. Partial recovery for top-K𝐾Kitalic_K ranking: Optimality of mle and suboptimality of the spectral method. The Annals of Statistics, 50 1618–1652.
  8. Robust dynamic assortment optimization in the presence of outlier customers. Operations Research.
  9. Dynamic assortment optimization with changing contextual information. The Journal of Machine Learning Research, 21 8918–8961.
  10. Spectral method and regularized mle are both optimal for top-K𝐾Kitalic_K ranking. Annals of statistics, 47 2204.
  11. Spectral mle: Top-K𝐾Kitalic_K rank aggregation from pairwise comparisons. In International Conference on Machine Learning. PMLR.
  12. Label ranking methods based on the Plackett-Luce model. In ICML.
  13. Central limit theorems and bootstrap in high dimensions. The Annals of Probability, 45 2309–2352.
  14. Improved central limit theorem and bootstrap approximations in high dimensions. arXiv preprint arXiv:1912.10529.
  15. Assortment optimization under variants of the nested logit model. Operations Research, 62 250–273.
  16. Rank aggregation methods for the web. In Proceedings of the 10th international conference on World Wide Web.
  17. Uncertainty quantification of mle for entity ranking with covariates. arXiv preprint arXiv:2212.09961.
  18. Ranking inferences based on the top choice of multiway comparisons. arXiv preprint arXiv:2211.11957.
  19. Constrained assortment optimization for the nested logit model. Management Science, 60 2583–2601.
  20. Uncertainty quantification in the Bradley-Terry-Luce model. arXiv preprint arXiv:2110.03874.
  21. Bayesian inference for Plackett-Luce ranking models. In proceedings of the 26th annual international conference on machine learning.
  22. Minimax-optimal inference from partial rankings. Advances in Neural Information Processing Systems, 27.
  23. A unified analysis of likelihood-based estimators in the Plackett–Luce model. arXiv preprint arXiv:2306.02821.
  24. Asymptotic theory of sparse Bradley–Terry model. The Annals of Applied Probability, 30 2491–2515.
  25. Hunter, D. R. (2004). MM algorithms for generalized Bradley-Terry models. The annals of statistics, 32 384–406.
  26. Top-K𝐾Kitalic_K rank aggregation from m𝑚mitalic_m-wise comparisons. IEEE Journal of Selected Topics in Signal Processing, 12 989–1004.
  27. Top-K𝐾Kitalic_K ranking from pairwise comparisons: When spectral ranking is optimal. arXiv preprint arXiv:1603.04153.
  28. Co-citation and co-authorship networks of statisticians. Journal of Business & Economic Statistics, 40 469–485.
  29. Meta-analysis on citations for statisticians. To Appear.
  30. Bayesian analysis of rank data with application to primate intelligence experiments. Journal of the American Statistical Association, 97 8–17.
  31. Estimating and exploiting the impact of photo layout: A structural approach. Available at SSRN 3470877.
  32. ℓ∞subscriptℓ\ell_{\infty}roman_ℓ start_POSTSUBSCRIPT ∞ end_POSTSUBSCRIPT-bounds of the mle in the btl model under general comparison graphs. In Uncertainty in Artificial Intelligence. PMLR.
  33. Lagrangian inference for ranking problems. Operations Research.
  34. Luce, R. D. (1959). Individual choice behavior: A theoretical analysis. John Wiley & Sons, Inc., New York; Chapman & Hall, Ltd., London.
  35. Massey, K. (1997). Statistical models applied to the rating of sports teams. Bluefield College, 1077.
  36. An empirical study of voting rules and manipulation with large datasets. Proceedings of COMSOC, 59.
  37. Preflib: A library for preferences http://www.preflib.org. In International conference on algorithmic decision theory. Springer.
  38. Fast and accurate inference of Plackett–Luce models. Advances in Neural Information Processing Systems, 28.
  39. Iterative ranking from pair-wise comparisons. Advances in Neural Information Processing Systems, 25.
  40. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35 27730–27744.
  41. Plackett, R. L. (1975). The analysis of permutations. Journal of the Royal Statistical Society: Series C (Applied Statistics), 24 193–202.
  42. Portnoy, S. (1986). On the central limit theorem in 𝐑psuperscript𝐑𝑝{\bf R}^{p}bold_R start_POSTSUPERSCRIPT italic_p end_POSTSUPERSCRIPT when p→∞→𝑝p\to\inftyitalic_p → ∞. Probab. Theory Related Fields, 73 571–583.
  43. Dynamic assortment optimization with a multinomial logit choice model and capacity constraint. Operations research, 58 1666–1680.
  44. Robust assortment optimization in revenue management under the multinomial logit choice model. Operations research, 60 865–882.
  45. Estimation from pairwise comparisons: Sharp minimax bounds with topology dependence. In Artificial intelligence and statistics. PMLR.
  46. Combinatorial inference on the optimal assortment in multinomial logit models. Available at SSRN 4371919.
  47. Asymptotics when the number of parameters tends to infinity in the Bradley-Terry model for paired comparisons. The Annals of Statistics, 27 1041–1060.
  48. Revenue-utility tradeoff in assortment optimization under the multinomial logit model with totally unimodular constraints. Management Science, 67 2845–2869.
  49. Online rank elicitation for Plackett-Luce: A dueling bandits approach. Advances in Neural Information Processing Systems, 28.
  50. Revenue management under a general discrete choice model of consumer behavior. Management Science, 50 15–33.
  51. Tropp, J. A. (2012). User-friendly tail bounds for sums of random matrices. Foundations of computational mathematics, 12 389–434.
  52. Bradley-Terry models in R: the BradleyTerry2 package. Journal of Statistical Software, 48 1–21.
  53. Estimating primary demand for substitutable products from sales transaction data. Operations Research, 60 313–334.
  54. Learning to rank with selection bias in personal search. In Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval.
  55. Assortment optimization under the paired combinatorial logit model. Operations Research, 68 741–761.
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets