Search

This study introduces a greedy algorithm for deriving decision rules from decision tree ensembles, targeting enhanced interpretability and generalization in distributed data environments. Decision rules, known for their transparency, provide an accessible method for knowledge extraction from data, facilitating decision-making processes across diverse fields. Traditional decision tree algorithms, such as CART and ID3, are employed to induce decision trees from bootstrapped datasets, which represent distributed data sources. Subsequently, a greedy algorithm is applied to derive decision rules that are true across multiple decision trees. Experiments are performed, taking into account knowledge representation and discovery perspectives. They show that, as the value of α, 0≤α<1, increases, shorter rules are obtained, and also it is possible to improve the classification accuracy of rule-based models.

Supervised discretisation is widely considered as far more advantageous than unsupervised transformation of attributes, because it helps to preserve the informative content of a variable, which is useful in classification. After discretisation, based on employed criteria, some attributes can be found irrelevant, and all their values can be represented in a discrete domain by a single interval. In consequence, such attributes are removed from considerations, and no knowledge is mined from them. The paper presents research focused on extended transformations of attribute values, thus combining supervised with unsupervised discretisation strategies. For all variables with single intervals returned from supervised algorithms, the ranges of values were transformed by unsupervised methods with varying numbers of bins. Resulting variants of the data were subjected to selected data mining techniques, and the performance of a group of classifiers was evaluated and compared. The experiments were performed on a stylometric task of authorship attribution.

The nature of the input features is one of the key factors indicating what kind of tools, methods, or approaches can be used in a knowledge discovery process. Depending on the characteristics of the available attributes, some techniques could lead to unsatisfactory performance or even may not proceed at all without additional preprocessing steps. The types of variables and their domains affect performance. Any changes to their form can influence it as well, or even enable some learners. On the other hand, the relevance of features for a task constitutes another element with a noticeable impact on data exploration. The importance of attributes can be estimated through the application of mechanisms belonging to the feature selection and reduction area, such as rankings. In the described research framework, the data form was conditioned on relevance by the proposed procedure of gradual discretisation controlled by a ranking of attributes. Supervised and unsupervised discretisation methods were employed to the datasets from the stylometric domain and the task of binary authorship attribution. For the selected classifiers, extensive tests were performed and they indicated many cases of enhanced prediction for partially discretised datasets.

In this article, we present a modification of the algorithm based on EAV (entity–attribute–value) model, for induction of decision rules, utilizing novel approach for attribute ranking. The selection of attributes used as premises of decision rules, is an important stage of the process of rules induction. In the presented approach, this task is realized using ranking of attributes based on standard deviation of attributes’ values per decision classes, which is considered as a distinguishability level. The presented approach allows to work not only with numerical values of attributes but also with categorical ones. For this purpose, an additional step of data transformation into a matrix format has been proposed. It allows to transform data table into a binary one with proper equivalents of categorical values of attributes and ensures independence of the influence of the attribute selection function from the data type of variables. The motivation for the proposed method is the development of an algorithm which allows to construct rules close to optimal ones in terms of length, while maintaining enough good classification quality. The experiments presented in the paper have been performed on data sets from UCI ML Repository, comparing results of the proposed approach with three selected greedy heuristics for induction of decision rules, taking into consideration classification accuracy and length and support of constructed rules. The obtained results show that for the most part of datasests, the average length of rules obtained for 80% of best attributes from the ranking is very close to values obtained for the whole set of attributes. In case of classification accuracy, for 50% of considered datasets, results obtained for 80% of best attributes from the ranking are higher or the same as results obtained for the whole set of attributes.

In this paper, we deal with distributed data represented either as a finite set T of decision tables with equal sets of attributes or a finite set I of information systems with equal sets of attributes. In the former case, we discuss a way to the study decision trees common to all tables from the set T: building a decision table in which the set of decision trees coincides with the set of decision trees common to all tables from T. We show when we can build such a decision table and how to build it in a polynomial time. If we have such a table, we can apply various decision tree learning algorithms to it. We extend the considered approach to the study of test (reducts) and decision rules common to all tables from T. In the latter case, we discuss a way to study the association rules common to all information systems from the set I: building a joint information system for which the set of true association rules that are realizable for a given row ρ and have a given attribute a on the right-hand side coincides with the set of association rules that are true for all information systems from I, have the attribute a on the right-hand side, and are realizable for the row ρ. We then show how to build a joint information system in a polynomial time. When we build such an information system, we can apply various association rule learning algorithms to it.

When patterns to be recognised are described by features of continuous type, discretisation becomes either an optional or necessary step in the initial data pre-processing stage. Characteristics of data, distribution of data points in the input space, can significantly influence the process of transformation from real-valued into nominal attributes, and the resulting performance of classification systems employing them. If data include several separate sets, their discretisation becomes more complex, as varying numbers of intervals and different ranges can be constructed for the same variables. The paper presents research on irregularities in data distribution, observed in the context of discretisation processes. Selected discretisation methods were used and their effect on the performance of decision algorithms, induced in classical rough set approach, was investigated. The studied input space was defined by measurable style-markers, which, exploited as characteristic features, facilitate treating a task of stylometric authorship attribution as classification

Conventional decision trees use queries each of which is based on one attribute. In this study, we also examine decision trees that handle additional queries based on hypotheses. This kind of query is similar to the equivalence queries considered in exact learning. Earlier, we designed dynamic programming algorithms for the computation of the minimum depth and the minimum number of internal nodes in decision trees that have hypotheses. Modification of these algorithms considered in the present paper permits us to build decision trees with hypotheses that are optimal relative to the depth or relative to the number of the internal nodes. We compare the length and coverage of decision rules extracted from optimal decision trees with hypotheses and decision rules extracted from optimal conventional decision trees to choose the ones that are preferable as a tool for the representation of information. To this end, we conduct computer experiments on various decision tables from the UCI Machine Learning Repository. In addition, we also consider decision tables for randomly generated Boolean functions. The collected results show that the decision rules derived from decision trees with hypotheses in many cases are better than the rules extracted from conventional decision trees.

Typically discretisation procedures are implemented as a part of initial pre-processing of data, before knowledge mining is employed. It means that conclusions and observations are based on reduced data, as usually by discretisation some information is discarded. The paper presents a different approach, with taking advantage of discretisation executed after data mining. In the described study firstly decision rules were induced from real-valued features. Secondly, data sets were discretised. Using categories found for attributes, in the third step conditions included in inferred rules were translated into discrete domain. The properties and performance of rule classifiers were tested in the domain of stylometric analysis of texts, where writing styles were defined through quantitative attributes of continuous nature. The performed experiments show that the proposed processing leads to sets of rules with significantly reduced sizes while maintaining quality of predictions, and allows to test many data discretisation methods at the acceptable computational costs.

In the paper, an approach for decision rules construction is proposed. It is studied from the point of view of the supervised machine learning task, i.e., classification, and from the point of view of knowledge representation. Generated rules provide comparable classification results to the dynamic programming approach for optimization of decision rules relative to length or support. However, the proposed algorithm is based on transformation of decision table into entity–attribute–value (EAV) format. Additionally, standard deviation function for computation of averages’ values of attributes in particular decision classes was introduced. It allows to select from the whole set of attributes only these which provide the highest degree of information about the decision. Construction of decision rules is performed based on idea of partitioning of a decision table into corresponding subtables. In opposite to dynamic programming approach, not all attributes need to be taken into account but only these with the highest values of standard deviation per decision classes. Consequently, the proposed solution is more time efficient because of lower computational complexity. In the framework of experimental results, support and length of decision rules were computed and compared with the values of optimal rules. The classification error for data sets from UCI Machine Learning Repository was also obtained and compared with the ones for dynamic programming approach. Performed experiments show that constructed rules are not far from the optimal ones and classification results are comparable to these obtained in the framework of the dynamic programming extension.