Search

Contrast pattern-based classifiers are an important family of both understandable and accurate classifiers. Nevertheless, these classifiers do not achieve good performance on class imbalance problems. In this paper, we introduce a new contrast pattern-based classifier for class imbalance problems. Our proposal for solving the class imbalance problem combines the support of the patterns with the class imbalance level at the classification stage of the classifier. From our experimental results, using highly imbalanced databases, we can conclude that our proposed classifier significantly outperforms the current contrast pattern-based classifiers designed for class imbalance problems. Additionally, we show that our classifier significantly outperforms other state-of-the-art classifiers not directly based on contrast patterns, which are also designed to deal with class imbalance problems.

An experimental study of the effect of class imbalance on quality measures for contrast patterns is presented.Our study was performed both regarding and disregarding the class imbalance level.The best quality measures for ranking patterns in class imbalance problems is provided.A guide for determining, according to the class imbalance ratio, the quality measure to apply, is provided. Contrast pattern-based classifiers rely on the discriminative power of contrast patterns. For this reason, many quality measures have been proposed to evaluate the quality of a contrast pattern. These measures allow to distinguish among contrast patterns with low and high discriminative ability for classification. In the literature, many comparative studies among quality measures for contrast patterns have been proposed but all of them were performed without taking into account the class imbalance level. However, in many class imbalance problems, those patterns extracted from the minority class have low support, which could negatively affect their discriminative ability. Therefore, in this paper, we present an experimental study of the effect of class imbalance on quality measures for contrast patterns. This study determines which quality measures for contrast patterns are the best for class imbalance problems; both regarding and disregarding the class imbalance level. Also, for the best quality measures we performed a pairwise comparison to determine which other quality measures have statistically similar behavior to them. This will help to simplify future research since it can be used only one quality measure among those with similar performance.

Automatic latent fingerprint identification is a useful tool for criminal investigation. However, the accuracy of identification reported in the state-of-the-art literature is low due to the distortion in latent fingerprint images. In this paper, we describe a new algorithm based on the use of clustering which is independent of the minutiae descriptors. The proposed technique improves the robustness of identification in the presence of large non-linear deformation which is associated with latent fingerprint images. The new algorithm finds multiple overlapping clusters of matching minutiae pairs which are merged together to find matching minutiae. Several experiments performed using latent fingerprint databases show that our proposed algorithm achieves higher accuracy than those presented in state-of-the-art literature. Moreover, the results show that the proposed algorithm is successful in dealing with the large distortion associated with latent fingerprints formed under the worst conditions.

The class imbalance problem is a challenge in supervised classification, since many classifiers are sensitive to class distribution, biasing their prediction towards the majority class. Usually, in imbalanced databases, contrast pattern miners extract a very large collection of patterns from the majority class but only a few patterns (or none) from the minority class. It causes that minority class objects have low support and they could be identified as noise and consequently discarded by the contrast pattern based classifier biasing the results towards the majority class. In the literature, the class imbalance problem is commonly faced by applying resampling methods. Therefore, in this paper, we present a study about the impact of using resampling methods for improving the performance of contrast pattern based classifiers in class imbalance problems. Experimental results using standard imbalanced databases show that there are statistically significant differences between using the classifier before and after applying resampling methods. Moreover, from this study, we provide a guide based on the class imbalance ratio for selecting a resampling method that jointly with a contrast pattern based classifier allows us to have good results in a class imbalance problem. HighlightsThe effect of resampling methods on contrast pattern based classifiers.A study of contrast pattern based classifiers in class imbalance problems.We show how improving the performance of contrast pattern based classifiers.We provide a rough guide for selecting the best resampling method regarding the IR.

Comparison of diversity generation procedures for mining contrast patterns.Diversity calculated based on the amount of total, unique, and minimal patterns.Three new deterministic methods for generating diversity in decision trees.Study of the influence of data type in diversity and accuracy of methods.Random Forest and Bagging are the best procedures. Most understandable classifiers are based on contrast patterns, which can be accurately mined from decision trees. Nevertheless, tree diversity must be ensured to mine a representative pattern collection. In this paper, we performed an experimental comparison among different diversity generation procedures. We compare diversity generated by each procedure based on the amount of total, unique, and minimal patterns extracted from the induced tree for different minimal support thresholds. This comparison, together with an accuracy and abstention experiment, shows that Random Forest and Bagging generate the most diverse and accurate pattern collection. Additionally, we study the influence of data type in the results, finding that Random Forest is best for categorical data and Bagging for numerical data. Comparison includes most known diversity generation procedures and three new deterministic procedures introduced here. These deterministic procedures outperform existing deterministic method, but are still outperformed by random procedures.

Pattern-based clustering algorithms return a set of patterns that describe the objects of each cluster. The most recent algorithms proposed in this approach extract patterns on numerical datasets by applying an a priori discretization process, which may cause information loss. In this paper, we introduce a new pattern-based clustering algorithm for numerical datasets, which does not need an a priori discretization on numerical features. The new algorithm extracts, from a collection of trees generated through a new induction procedure, a small subset of patterns useful for clustering. Experimental results show that the patterns extracted by the proposed algorithm allows to build a pattern-based clustering algorithm, which obtains better clustering results than recent pattern-based clustering algorithms. In addition, the proposed algorithm obtains similar clustering results, in quality, than traditional clustering algorithms.

In this paper, we introduce an efficient algorithm for mining discriminative regularities on databases with mixed and incomplete data. Unlike previous methods, our algorithm does not apply an a priori discretization on numerical features; it extracts regularities from a set of diverse decision trees, induced with a special procedure. Experimental results show that a classifier based on the regularities obtained by our algorithm attains higher classification accuracy, using fewer discriminative regularities than those obtained by previous pattern-based classifiers. Additionally, we show that our classifier is competitive with traditional and state-of-the-art classifiers.