@conference {6007647, title = {Using the Adaboost algorithm for extracting fuzzy rules from low quality data: Some preliminary results}, booktitle = {2011 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2011)}, year = {2011}, month = {June}, pages = {1263-1270}, abstract = {When the Adaboost algorithm is used for extracting fuzzy rules from data, each rule is regarded as a weak learner, and knowledge bases as assimilated to ensembles. In this paper we propose an extension of this framework for obtaining fuzzy rule-based classifiers from imprecise data. In the new approach, the mentioned search of the best rule at each iteration is carried out by a genetic algorithm with a fuzzy fitness function. The instances will be assigned fuzzy weights, however each fuzzy rule will be associated to a crisp number of votes.}, keywords = {Adaboost algorithm, Boosting, data handling, Electronic mail, fuzzy fitness function, fuzzy rule extraction, fuzzy rule-based classifiers, fuzzy set theory, Fuzzy systems, fuzzy weights, genetic algorithm, genetic algorithms, Genetic Fuzzy Systems, knowledge based systems, knowledge bases, learning (artificial intelligence), Low Quality Data, Merging, Optimization, pattern classification, Training}, issn = {1098-7584}, doi = {10.1109/FUZZY.2011.6007647}, author = {A. M. Palacios and L. S{\'a}nchez and I. Couso} } @conference {5584797, title = {Preprocessing vague imbalanced datasets and its use in genetic fuzzy classifiers}, booktitle = {International Conference on Fuzzy Systems}, year = {2010}, month = {July}, pages = {1-8}, abstract = {When there is a substantial difference between the number of cases of the majority and minority classes, minimum error-based classification systems tend to overlook these last instances. This can be corrected either by preprocessing the dataset or by altering the objective function of the classifier. In this paper we analyze the first approach, in the context of genetic fuzzy systems (GFS), and in particular of those that can operate with imprecisely observed and low quality data. We will analyze the different preprocessing mechanisms of imbalanced datasets and will show the necessity of extending these for solving those problems where the data is both imprecise and im-balanced. In addition, we include a comprehensive description of a new algorithm, able to preprocess imprecise imbalanced datasets. Several real-world datasets are used to evaluate the proposal.}, keywords = {Classification algorithms, Context, data handling, Euclidean distance, fuzzy set theory, Fuzzy systems, genetic algorithms, genetic fuzzy classifier, genetic fuzzy system, Genetics, imbalanced dataset preprocessing, minimum error based classification system, Nearest neighbor searches, objective function, pattern classification, Pediatrics, Training}, issn = {1098-7584}, doi = {10.1109/FUZZY.2010.5584797}, author = {A. M. Palacios and L. S{\'a}nchez and I. Couso} } @article {4286977, title = {Advocating the Use of Imprecisely Observed Data in Genetic Fuzzy Systems}, journal = {IEEE Transactions on Fuzzy Systems}, volume = {15}, number = {4}, year = {2007}, month = {Aug}, pages = {551-562}, abstract = {In our opinion, and in accordance with current literature, the precise contribution of genetic fuzzy systems to the corpus of the machine learning theory has not been clearly stated yet. In particular, we question the existence of a set of problems for which the use of fuzzy rules, in combination with genetic algorithms, produces more robust models, or classifiers that are inherently better than those arising from the Bayesian point of view. We will show that this set of problems actually exists, and comprises interval and fuzzy valued datasets, but it is not being exploited. Current genetic fuzzy classifiers deal with crisp classification problems, where the role of fuzzy sets is reduced to give a parametric definition of a set of discriminant functions, with a convenient linguistic interpretation. Provided that the customary use of fuzzy sets in statistics is vague data, we propose to test genetic fuzzy classifiers over imprecisely measured data and design experiments well suited to these problems. The same can be said about genetic fuzzy models: the use of a scalar fitness function assumes crisp data, where fuzzy models, a priori, do not have advantages over statistical regression.}, keywords = {Bayesian methods, data handling, Design for experiments, discriminant function, fuzzy fitness function, fuzzy rule-based classifiers, fuzzy rule-based models, fuzzy rules, fuzzy set theory, Fuzzy sets, Fuzzy systems, genetic algorithm, genetic algorithms, genetic fuzzy classifiers, Genetic Fuzzy Systems, learning (artificial intelligence), linguistic interpretation, machine learning, machine learning theory, Parametric statistics, pattern classification, Robustness, scalar fitness function, statistical analysis, Stochastic resonance, vague data}, issn = {1063-6706}, doi = {10.1109/TFUZZ.2007.895942}, author = {L. S{\'a}nchez and I. Couso} }