Apport de la négation pour la classification supervisée à l'aide d'associations

L'utilisation de regles d'association est une methode bien connue pour des tâches de classification supervisee. Traditionnellement, les regles utilisees sont de la forme X -> c, ou X est un ensemble d'attributs et c est une valeur de classe. Nous nous interessons a la generalisation naturelle de ces regles consistant a autoriser des attributs negatifs en premisse ainsi que des valeurs de classe negatives en conclusion. Comprendre l'impact des regles avec negation dans un processus de classification est une tâche cruciale. Nous proposons un classifieur utilisant de telles regles et etudions l'apport de ces dernieres pour la tâche de classification. Contrairement a l'intuition, il s'avere que, toutes choses egales par ailleurs, l'utilisation de ces regles avec negation en classification est delicate.

[1]  Elena Baralis,et al.  Majority Classification by Means of Association Rules , 2003, PKDD.

[2]  William W. Cohen.  Fast Effective Rule Induction , 1995, ICML.

[3]  Jean-François Boulicaut,et al.  Towards the Tractable Discovery of Association Rules with Negations , 2000, FQAS.

[4]  Vladimir Gurvich,et al.  An Efficient Implementation of a Quasi-polynomial Algorithm for Generating Hypergraph Transversals , 2003, ESA.

[5]  Jaideep Srivastava,et al.  Selecting the right interestingness measure for association patterns , 2002, KDD.

[6]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[7]  Osmar R. Zaïane,et al.  On Pruning and Tuning Rules for Associative Classifiers , 2005, KES.

[8]  Leonid Khachiyan,et al.  On the Complexity of Dualization of Monotone Disjunctive Normal Forms , 1996, J. Algorithms.

[9]  Elena Baralis,et al.  Associative text categorization exploiting negated words , 2006, SAC '06.

[10]  Jianyong Wang,et al.  On Mining Instance-Centric Classification Rules , 2006, IEEE Transactions on Knowledge and Data Engineering.

[11]  Mohammed J. Zaki.  Generating non-redundant association rules , 2000, KDD '00.

[12]  Jean-François Boulicaut,et al.  A Survey on Condensed Representations for Frequent Sets , 2004, Constraint-Based Mining and Inductive Databases.

[13]  Elena Baralis,et al.  A lazy approach to pruning classification rules , 2002, 2002 IEEE International Conference on Data Mining, 2002. Proceedings..

[14]  Catherine Blake,et al.  UCI Repository of machine learning databases , 1998 .

[15]  R. Mike Cameron-Jones,et al.  FOIL: A Midterm Report , 1993, ECML.

[16]  Wynne Hsu,et al.  Integrating Classification and Association Rule Mining , 1998, KDD.

[17]  Samir Elloumi,et al.  Integrated Generic Association Rule Based Classifier , 2007, 18th International Workshop on Database and Expert Systems Applications (DEXA 2007).

[18]  Rajeev Motwani,et al.  Beyond market baskets: generalizing association rules to correlations , 1997, SIGMOD '97.

[19]  Yann Mathet,et al.  Classification de textes d'opinions : une approche mixte n-grammes et sémantique , 2007 .

[20]  R. Agarwal.  Fast Algorithms for Mining Association Rules , 1994, VLDB 1994.

[21]  Toon Calders,et al.  Minimal k-Free Representations of Frequent Sets , 2003, PKDD.

[22]  Patrick Meyer,et al.  Mesurer l'intérêt des règles d'association , 2005, EGC.

[23]  Jian Pei,et al.  CMAR: accurate and efficient classification based on multiple class-association rules , 2001, Proceedings 2001 IEEE International Conference on Data Mining.

[24]  Heikki Mannila,et al.  Levelwise Search and Borders of Theories in Knowledge Discovery , 1997, Data Mining and Knowledge Discovery.

[25]  Gerd Stumme,et al.  Generating a Condensed Representation for Association Rules , 2005, Journal of Intelligent Information Systems.

[26]  Osmar R. Zaïane,et al.  An associative classifier based on positive and negative rules , 2004, DMKD '04.

[27]  Davy Janssens,et al.  Improving Associative Classification by Incorporating Novel Interestingness Measures , 2005, ICEBE.

[28]  Jiawei Han,et al.  CPAR: Classification based on Predictive Association Rules , 2003, SDM.