• Corpus ID: 49667363

Towards Non-Parametric Learning to Rank

@article{Liu2018TowardsNL,
  title={Towards Non-Parametric Learning to Rank},
  author={Ao Liu and Qiong Wu and Zhenming Liu and Lirong Xia},
  journal={ArXiv},
  year={2018},
  volume={abs/1807.03395}
}
This paper studies a stylized, yet natural, learning-to-rank problem and points out the critical incorrectness of a widely used nearest neighbor algorithm. We consider a model with $n$ agents (users) $\{x_i\}_{i \in [n]}$ and $m$ alternatives (items) $\{y_j\}_{j \in [m]}$, each of which is associated with a latent feature vector. Agents rank items nondeterministically according to the Plackett-Luce model, where the higher the utility of an item to the agent, the more likely this item will be… 
View PDF on arXiv

Figures from this paper

References

SHOWING 1-10 OF 38 REFERENCES

Nonparametric Preference Completion

This work proposes a $k-nearest neighbors-like algorithm and proves that it is consistent, which is the first consistency result for the collaborative preference completion problem in a nonparametric setting.

Blind Regression: Nonparametric Regression for Latent Variable Models via Collaborative Filtering

Inspired by the classical Taylor's expansion for differentiable functions, a prediction algorithm that is consistent for all Lipschitz functions is provided, and it is proved that the expected fraction of estimates with error greater than $\epsilon$ is less than the variance of the additive entry-wise noise term.

Preference Completion from Partial Rankings

A novel and efficient algorithm for the collaborative preference completion problem, which involves jointly estimating individualized rankings for a set of entities over a shared set of items, based on a limited number of observed affinity values, that can fit any consistent partial ranking over a subset of the items represented as a directed acyclic graph.

Computing Parametric Ranking Models via Rank-Breaking

This work characterize the breakings for which the estimator is consistent for random utility models (RUMs) including Plackett-Luce and Normal-RUM, develop a general sufficient condition for a full breaking to be the only consistent breaking, and provide a trichotomy theorem in regard to single-edge breakings.

Preference Completion: Large-scale Collaborative Ranking from Pairwise Comparisons

A large-scale non-convex implementation of AltSVM is developed that trains a factored form of the matrix via alternating minimization, and scales and parallelizes very well to large problem settings.

Learning to Recommend Accurate and Diverse Items

A diversified collaborative filtering algorithm (DCF) is proposed to solve the coupled problems of parameterized matrix factorization and structural learning, and a new pairwise accuracy metric and a normalized topic coverage diversity metric to measure the performance of accuracy and diversity respectively.

Bayesian Plackett–Luce Mixture Models for Partially Ranked Data

A Bayesian finite mixture of Plackett–Luce models to account for unobserved sample heterogeneity of partially ranked data is introduced and an efficient way to incorporate the latent group structure in the data augmentation approach and the derivation of existing maximum likelihood procedures as special instances of the proposed Bayesian method is described.

Rank Centrality: Ranking from Pairwise Comparisons

Experimental evaluations on synthetic data sets generated according to the popular Bradley-Terry-Luce (BTL) model show that the proposed Rank Centrality algorithm performs as well as the maximum likelihood estimator for that model and outperforms other popular ranking algorithms.

Data-driven Rank Breaking for Efficient Rank Aggregation

This paper provides the optimal rank-breaking estimator, which not only achieves consistency but also achieves the best error bound, which allows the fundamental tradeoff between accuracy and complexity to be characterized.

Label Ranking Methods based on the Plackett-Luce Model

This paper introduces two new methods for label ranking based on a probabilistic model of ranking data, called the Plackett-Luce model, which estimates a global model in which the PL parameters are represented as functions of the instance.