{rfName}
Re

Indexat a

Llicència i ús

Altmetrics

Grant support

This work was supported in part by the funding for the ByoPiC Project from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program under Grant ERC-2015-AdG 695561. The work of A.Decelle was supported by the Comunidad de Madrid and the Complutense University of Madrid (Spain) through the Atraccion de Talento Program under Grant 2019-T1/TIC-13298.

Anàlisi d'autories institucional

Decelle, AurelienAutor o coautor
Compartir
Publicacions
>
Article

Regularization of Mixture Models for Robust Principal Graph Learning

Publicat a:Ieee Transactions On Pattern Analysis And Machine Intelligence. 44 (12): 9119-9130 - 2022-12-01 44(12), DOI: 10.1109/TPAMI.2021.3124973

Autors: Bonnaire, Tony; Decelle, Aurelien; Aghanim, Nabila

Afiliacions

CNRS, Inst Astrophys Spatiale, F-91405 Bures Sur Yvette, France - Autor o coautor
CNRS, Lab Rech Informat, F-91190 Gif Sur Yvette, France - Autor o coautor
Univ Complutense, Dept Fis Teor 1, Madrid 28040, Spain - Autor o coautor
Univ Paris Saclay, F-91190 Gif Sur Yvette, France - Autor o coautor

Resum

A regularized version of Mixture Models is proposed to learn a principal graph from a distribution of D-dimensional datapoints. In the particular case of manifold learning for ridge detection, we assume that the underlying structure can be modeled as a graph acting like a topological prior for the Gaussian clusters turning the problem into a maximum a posteriori estimation. Parameters of the model are iteratively estimated through an Expectation-Maximization procedure making the learning of the structure computationally efficient with guaranteed convergence for any graph prior in a polynomial time. We also embed in the formalism a natural way to make the algorithm robust to outliers of the pattern and heteroscedasticity of the manifold sampling coherently with the graph structure. The method uses a graph prior given by the minimum spanning tree that we extend using random sub-samplings of the dataset to take into account cycles that can be observed in the spatial distribution.

Paraules clau
Dimensionality reductionExpectation-maximizationGaussian mixture modelsGraph regularizationLaplacian eigenmapManifold learninPrincipal graph

Indicis de qualitat

Impacte bibliomètric. Anàlisi de la contribució i canal de difusió

El treball ha estat publicat a la revista Ieee Transactions On Pattern Analysis And Machine Intelligence a causa de la seva progressió i el bon impacte que ha aconseguit en els últims anys, segons l'agència WoS (JCR), s'ha convertit en una referència en el seu camp. A l'any de publicació del treball, 2022, es trobava a la posició 2/145, aconseguint així situar-se com a revista Q1 (Primer Cuartil), en la categoria Computer Science, Artificial Intelligence. Destacable, igualment, el fet que la revista està posicionada per sobre del Percentil 90.

Des d'una perspectiva relativa, i atenent a l'indicador de impacte normalitzat calculat a partir del Field Citation Ratio (FCR) de la font Dimensions, proporciona un valor de: 2.88, el que indica que, comparat amb treballs en la mateixa disciplina i en el mateix any de publicació, el situa com un treball citat per sobre de la mitjana. (font consultada: Dimensions May 2025)

Concretament, i atenent a les diferents agències d'indexació, aquest treball ha acumulat, fins a la data 2025-05-25, el següent nombre de cites:

  • WoS: 3
  • Scopus: 6
  • OpenCitations: 4
Impacte i visibilitat social

Des de la dimensió d'influència o adopció social, i prenent com a base les mètriques associades a les mencions i interaccions proporcionades per agències especialitzades en el càlcul de les denominades "Mètriques Alternatives o Socials", podem destacar a data 2025-05-25:

  • L'ús, des de l'àmbit acadèmic evidenciat per l'indicador de l'agència Altmetric referit com a agregacions realitzades pel gestor bibliogràfic personal Mendeley, ens dona un total de: 9.
  • L'ús d'aquesta aportació en marcadors, bifurcacions de codi, afegits a llistes de favorits per a una lectura recurrent, així com visualitzacions generals, indica que algú està fent servir la publicació com a base del seu treball actual. Això pot ser un indicador destacat de futures cites més formals i acadèmiques. Aquesta afirmació està avalada pel resultat de l'indicador "Capture", que aporta un total de: 6 (PlumX).

Amb una intenció més de divulgació i orientada a audiències més generals, podem observar altres puntuacions més globals com:

  • El Puntuació total de Altmetric: 4.75.
  • El nombre de mencions a la xarxa social X (abans Twitter): 9 (Altmetric).
Anàlisi del lideratge dels autors institucionals

Aquest treball s'ha realitzat amb col·laboració internacional, concretament amb investigadors de: France.