<b>PerMallows</b>: An<i>R</i>Package for Mallows and Generalized Mallows Models

Irurozki, Ekhiñe; Calvo, B.; Lozano, José A.

doi:10.18637/jss.v071.i12

Cited by 44 publications

(42 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The fact that recursive descriptions of combinatorial objects can be translated into algorithms of random generation is classical [18]. Several generators can also be found in the literature [26]. We refer the reader interested on more details to [25].…”

Section: Results On Enumerative Combinatoricsmentioning

confidence: 99%

See 1 more Smart Citation

Mallows and generalized Mallows model for matchings

Irurozki¹,

Calvo²,

Lozano³

2019

Bernoulli

Self Cite

View full text Add to dashboard Cite

The Mallows and Generalized Mallows Models are two of the most popular probability models for distributions on permutations. In this paper we consider both models under the Hamming distance. This models can be seen as models for matchings instead of models for rankings. These models can not be factorized, which contrasts with the popular MM and GMM under Kendall's-τ and Cayley distances. In order to overcome the computational issues that the models involve, we introduce a novel method for computing the partition function. By adapting this method we can compute the expectation, joint and conditional probabilities. All these methods are the basis for three sampling algorithms, which we propose and analyze. Moreover, we also propose a learning algorithm. All the algorithms are analyzed both theoretically and empirically, using synthetic and real data from the context of e-learning and Massive Open Online Courses (MOOC).

show abstract

Section: Results On Enumerative Combinatoricsmentioning

confidence: 99%

“…The code used to run the experiments has been made public in the CRAN repository by the name of PerMallows. A manuscript introducing it can be found in [26]. Moreover, Appendix E shows how easy it is to fit and sample distributions a GMM with it.…”

Section: Methodsmentioning

confidence: 99%

Mallows and generalized Mallows model for matchings

Irurozki¹,

Calvo²,

Lozano³

2019

Bernoulli

Self Cite

View full text Add to dashboard Cite

show abstract

“…We use the methods described in [8] to obtain these uniformly at random solutions π i j for the different distances. In order to estimate the attraction basin size of π * , we proceed in a similar way to the previous method but, we work with the different subsets D i independently.…”

Section: Distance-based Methods (Dm)mentioning

confidence: 99%

Estimating Attraction Basin Sizes

Hernando

Mendiburu

Lozano

2016

Advances in Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

Abstract. The performance of local search algorithms is influenced by the properties that the neighborhood imposes on the search space. Among these properties, the number of local optima has been traditionally considered as a complexity measure of the instance, and different methods for its estimation have been developed. The accuracy of these estimators depends on properties such as the relative attraction basin sizes. As calculating the exact attraction basin sizes becomes unaffordable for moderate problem sizes, their estimations are required. The lack of techniques achieving this purpose encourages us to propose two methods that estimate the attraction basin size of a given local optimum. The first method takes uniformly at random solutions from the whole search space, while the second one takes into account the structure defined by the neighborhood. They are tested on different instances of problems in the permutation space, considering the swap and the adjacent swap neighborhoods.

show abstract

“…In the simulation study, ranking data were generated according to a Mallows model (Irurozki, Calvo, & Lozano, ), which is an exponential model defined by a central permutation α and a spread (or dispersion) parameter θ . When θ >0 (with θ =0 we obtain the uniform distribution), α is the mode of the distribution, that is, the permutation with the highest probability.…”

Section: Experimental Evaluationmentioning

confidence: 99%

A new position weight correlation coefficient for consensus ranking process without ties

Plaia

Buscemi

Sciandra

2019

Stat

View full text Add to dashboard Cite

Preference data represent a particular type of ranking data where a group of people give their preferences over a set of alternatives. The traditional metrics between rankings do not take into account the importance of swapping elements similar among them (element weights) or elements belonging to the top (or to the bottom) of an ordering (position weights). Following the structure of the τx proposed by Emond and Mason and the class of weighted Kemeny–Snell distances, a proper rank correlation coefficient is defined for measuring the correlation among weighted position rankings without ties. The one‐to‐one correspondence between the weighted distance and the rank correlation coefficient holds, analytically speaking, using both equal and decreasing weights. In order to determine the consensus ranking among rankings, related to a set of subjects, the new coefficient is maximized modifying suitably a branch‐and‐bound algorithm proposed in literature.

show abstract

PerMallows: AnRPackage for Mallows and Generalized Mallows Models

Cited by 44 publications

References 18 publications

Mallows and generalized Mallows model for matchings

Mallows and generalized Mallows model for matchings

Estimating Attraction Basin Sizes

A new position weight correlation coefficient for consensus ranking process without ties

Contact Info

Product

Resources

About