Jerzy Błaszczyński

Sequential covering rule induction algorithm for variable consistency rough set approaches

We present a general rule induction algorithm based on sequential covering, suitable for variable consistency rough set approaches. This algorithm, called VC-DomLEM, can be used for both ordered and non-ordered data. In the case of ordered data, the rough set model employs dominance relation, and in the case of non-ordered data, it employs indiscernibility relation. VC-DomLEM generates a minimal set of decision rules. These rules are characterized by a satisfactory value of the chosen consistency measure. We analyze properties of induced decision rules, and discuss conditions of correct rule induction. Moreover, we show how to improve rule induction efficiency due to application of consistency measures with desirable monotonicity properties.

Monotonic variable consistency rough set approaches

We consider probabilistic rough set approaches based on different versions of the definition of rough approximation of a set. In these versions, consistency measures are used to control assignment of objects to lower and upper approximations. Inspired by some basic properties of rough sets, we find it reasonable to require from these measures several properties of monotonicity. We consider three types of monotonicity properties: monotonicity with respect to the set of attributes, monotonicity with respect to the set of objects, and monotonicity with respect to the dominance relation. We show that consistency measures used so far in the definition of rough approximation lack some of these monotonicity properties. This observation led us to propose new measures within two kinds of rough set approaches: Variable Consistency Indiscernibility-based Rough Set Approaches (VC-IRSA) and Variable Consistency Dominance-based Rough Set Approaches (VC-DRSA). We investigate properties of these approaches and compare them to previously proposed Variable Precision Rough Set (VPRS) model, Rough Bayesian (RB) model, and previous versions of VC-DRSA.

Incremental Induction of Decision Rules from Dominance-based Rough Approximations

Abstract An incremental algorithm generating satisfactory decision rules and a rule post-processing technique are presented. The rule induction algorithm is based on the Apriori algorithm. It is extended to handle preference-ordered domains of attributes (called criteria) within Variable Consistency Dominance-based Rough Set Approach. It deals, moreover, with the problem of missing values in the data set. The algorithm has been designed for medical applications which require: (i) a careful selection of the set of decision rules representing medical experience and (ii) an easy update of these decision rules because of data set evolving in time, and (iii) not only a high predictive capacity of the set of decision rules but also a thorough explanation of a proposed decision. To satisfy all these requirements, we propose an incremental algorithm for induction of a satisfactory set of decision rules and a post-processing technique on the generated set of rules. Userʼns preferences with respect to attributes are also taken into account. A measure of the quality of a decision rule is proposed. It is used to select the most interesting representatives in the final set of rules.

Neighbourhood sampling in bagging for imbalanced data

Abstract Various approaches to extend bagging ensembles for class imbalanced data are considered. First, we review known extensions and compare them in a comprehensive experimental study. The results show that integrating bagging with under-sampling is more powerful than over-sampling. They also allow to distinguish Roughly Balanced Bagging as the most accurate extension. Then, we point out that complex and difficult distribution of the minority class can be handled by analyzing the content of a neighbourhood of examples. In our study we show that taking into account such local characteristics of the minority class distribution can be useful both for analyzing performance of ensembles with respect to data difficulty factors and for proposing new generalizations of bagging. We demonstrate it by proposing Neighbourhood Balanced Bagging, where sampling probabilities of examples are modified according to the class distribution in their neighbourhood. Two of its versions are considered: the first one keeping a larger size of bootstrap samples by hybrid over-sampling and the other reducing this size with stronger under-sampling. Experiments prove that the first version is significantly better than existing over-sampling bagging extensions while the other version is competitive to Roughly Balanced Bagging. Finally, we demonstrate that detecting types of minority examples depending on their neighbourhood may help explain why some ensembles work better for imbalanced data than others.

Integrating selective pre-processing of imbalanced data with Ivotes ensemble

In the paper we present a new framework for improving classifiers learned from imbalanced data. This framework integrates the SPIDER method for selective data pre-processing with the Ivotes ensemble. The goal of such integration is to obtain improved balance between the sensitivity and specificity for the minority class in comparison to a single classifier combined with SPIDER, and to keep overall accuracy on a similar level. The IIvotes framework was evaluated in a series of experiments, in which we tested its performance with two types of component classifiers (tree- and rule-based). The results show that IIvotes improves evaluation measures. They demonstrated advantages of the abstaining mechanism (i.e., refraining from predictions by component classifiers) in IIvotes rule ensembles.

A green chemistry-based classification model for the synthesis of silver nanoparticles

The assessment of the implementation of green chemistry principles in the syntheses of nanomaterials is a complex decision-making problem that necessitates the integration of several evaluation criteria. Multiple Criteria Decision Aiding (MCDA) provides support for such a challenge. One of its methods – Dominance-based Rough Set Approach (DRSA) – was used in this research to develop a model for the green chemistry-based classification of silver nanoparticle synthesis protocols into preference-ordered performance classes. DRSA allowed integration of knowledge from both peer-reviewed literature and experts (decision makers, DMs) in the field, resulting in a model composed of decision rules that are logical statements in the form: “if conditions, then decision”. The approach provides the basis for the design of rules for the greener synthesis of silver nanoparticles. Decision rules are supported by synthesis protocols that enforce the principles of green chemistry to various extents, resulting in robust recommendations for the development and assessment of silver nanoparticle synthesis that perform at one of five pre-determined levels. The DRSA-based approach is transparent and structured and can be easily updated. New perspectives and criteria could be added into the model if relevant data were available and domain-specific experts could collaborate through the MCDA procedure.

jMAF - Dominance-Based Rough Set Data Analysis Framework

We present a rough set data analysis software jMAF. It employs java Rough Set (jRS) library in which are implemented data analysis methods provided by the (variable consistency) Dominance-based Rough Set Approach (DRSA). The chapter also provides some basics of the DRSA and of its variable consistency extension.

Extending Bagging for Imbalanced Data

Various modifications of bagging for class imbalanced data are discussed. An experimental comparison of known bagging modifications shows that integrating with undersampling is more powerful than oversampling. We introduce Local-and-Over-All Balanced bagging where probability of sampling an example is tuned according to the class distribution inside its neighbourhood. Experiments indicate that this proposal is competitive to best undersampling bagging extensions.

Variable consistency bagging ensembles

In this paper we claim that the classification performance of bagging classifier can be improved by drawing to bootstrap samples objects being more consistent with their assignment to decision classes. We propose a variable consistency generalization of the bagging scheme where such sampling is controlled by two types of measures of consistency: rough membership and monotonic e measure. The usefulness of this proposal is experimentally confirmed with various rule and tree base classifiers. The results of experiments show that variable consistency bagging improves classification accuracy on inconsistent data.

On variable consistency dominance-based rough set approaches

We consider different variants of Variable Consistency Dominance-based Rough Set Approach (VC-DRSA). These variants produce more general (extended) lower approximations than those computed by Dominance-based Rough Set Approach (DRSA), (i.e., lower approximations that are supersets of those computed by DRSA). They define lower approximations that contain objects characterized by a strong but not necessarily certain relation with approximated sets. This is achieved by introduction of parameters that control consistency of objects included in lower approximations. We show that lower approximations generalized in this way enable us to observe dependencies that remain undiscovered by DRSA. Extended lower approximations are also a better basis for rule generation. In the paper, we focus our considerations on different definitions of generalized lower approximations. We also show definitions of VC-DRSA decision rules, as well as their application to classification/sorting and ranking/choice problems.